Cosmetic Preservatives Explained: How Formulas Stay Protected Through Storage and Use

A jar of face cream is opened, touched, closed, carried into a steamy bathroom and opened again, often every day for months. Throughout that time it is expected to stay safe and to behave the way it did on the first day. Preservatives are part of how that happens, but they are only one part. This guide explains the rest: where microorganisms come from, what else can go wrong with a formula, how pH, water and packaging change the picture, what the main preservative families are, what "natural" and "preservative-free" labels can and cannot tell you, and how shelf life and the open-jar symbol are supposed to be decided.

Quick answer

• Cosmetic preservatives are ingredients that prevent or reduce the growth of bacteria and fungi in a product during storage and consumer use.
• Protection comes from the whole product: the formula, raw-material quality, manufacturing hygiene, the pack, and how people actually use it. A preservative ingredient cannot make up for weaknesses in the rest.
• Water-rich products such as creams, lotions and liquid eye products usually need deliberate preservation. Low-water and anhydrous products can be lower-risk, but that has to be assessed, not assumed.
• "Natural", "preservative-free" or "airless" each describe one feature of a product. None of them is a safety result on its own.
• Shelf life and period-after-opening (PAO) figures are evidence questions for the safety assessment, not numbers chosen for the label. In the EU, the PAO symbol is required where minimum durability exceeds 30 months, except where an after-opening period is not relevant to the product.

Start Here: What Cosmetic Preservation Is

Preservative ingredient versus complete preservation system

A preservative is an ingredient. The European Commission describes it as "a natural or synthetic ingredient that is added to products to prevent them from spoiling" (European Commission, preservatives). The same page explains why the ingredient matters: without preservation, products handled directly by consumers can become contaminated, leading to spoilage, loss of performance and possibly irritation or infection.

A preservation system is something larger. The International Cooperation on Cosmetic Regulation (ICCR), a group of regulators and industry associations, describes it in its preservatives FAQs as the combination of a product's physical properties, its packaging and the preservative ingredients (ICCR FAQs, PDF). This guide follows that definition: whether a product is "well preserved" depends on the system and the evidence for that product, not just on which preservative it contains.

Who this guide is for

The guide is written for three kinds of readers. Consumers can use it to understand labels, storage advice and why a cream that looks fine may still be past its best. Product developers can use it as a shared vocabulary for conversations with formulators, laboratories and safety assessors. Brand owners can use it to see which preservation questions belong in a product brief before samples are made. It does not give formulas, dose levels or laboratory methods, which belong with qualified professionals.

How Microorganisms Reach Cosmetic Products

Bacteria and fungi live on skin, in air and in water, and a product meets them many times between the raw-material drum and the last use. The US Food and Drug Administration (FDA) lists the main routes: contaminated raw materials, water or other ingredients; an ineffective preservation system; packaging that does not adequately protect the product; poor shipping or storage conditions; and consumer use, such as dipping fingers into the product (FDA, Microbiological Safety and Cosmetics).

It helps to group these routes by stage, because each stage is controlled by different people.

Raw materials, water and manufacturing

The first opportunity is before mixing begins. Many raw materials are not sterile. Botanical extracts, mineral powders, gums and proteins can carry microorganisms from where they were grown, mined or processed.

Water is often the largest ingredient in a cream or lotion, so its quality matters. So do the surfaces it touches: tanks, pipes, hoses, filling nozzles, and the people operating them.

Packaging, storage, distribution and repeated use

Containers and closures can carry contamination if they are not stored properly before filling; afterwards, the pack is the main barrier. Heat in transit or storage also matters. FDA advises consumers not to leave cosmetics somewhere hot, such as a car, because heat can break preservatives down and help bacteria and fungi grow faster (FDA, Shelf Life and Expiration Dating).

The longest exposure usually comes during use. A face cream in an open jar is touched with fingertips at each application. Shampoo can take in water when the bottle is opened in the shower. Mascara is used right next to the eye. FDA lists habits that make this worse: sharing products, adding water or saliva to dried mascara, and using testers at cosmetic counters, which FDA says are even more likely to become contaminated than the same products at home.

One study shows how much contamination real use adds. Bashir and Lambert examined used lipsticks, lip glosses, eyeliners, mascaras and make-up sponges donated by consumers in the UK. Between 79% and 90% of the used products contained bacteria. The sponges carried far higher loads than the other products, and the study notes that 93% of them had never been cleaned (Bashir and Lambert, 2020, abstract). The study describes products after use, which is the point: use is a contamination route in its own right.

Bacteria, yeasts and moulds

The microorganisms involved fall into two broad groups. Bacteria are single-celled organisms; some grow quickly in water-rich products. Fungi include yeasts, which are also single-celled, and moulds, which form threads and spores and can grow visibly on surfaces. A preservative that works well against one group may do little against the other, which is one reason preservative systems are often combinations.

Organisms reported in contaminated cosmetics include Pseudomonas aeruginosa and Staphylococcus aureus, which a review in Contact Dermatitis describes as frequently found (Lundov et al., 2009, abstract). The used-product study above also found S. aureus and E. coli. FDA states that injuries from contaminated cosmetics are not common, but they can be serious. It cites serious infections linked to contaminated tattoo inks, eye-area products, and lotions and mouthwashes used in hospitals.

This guide stops at that level. Growing, identifying or deliberately introducing microorganisms is laboratory work for trained staff with validated methods, and nothing here is an instruction for it.

Microbial Spoilage, Oxidation and Physical Instability

When a product "goes off", people often blame the preservative. In practice, a cosmetic can fail in at least three different ways. Each has its own causes, protective ingredients and evidence, and mixing them up leads to the wrong fix.

Three different failure mechanisms

Microbial spoilage is the growth of bacteria, yeasts or moulds in the product. Microorganisms can feed on ingredients, break them down and change the product's properties, and some may harm the user. Spoilage can sometimes be seen or smelled, but not always.

Oxidation is a chemical change, usually triggered by oxygen and often sped up by light, heat or trace metals. Plant oils and other unsaturated lipids are especially exposed. Oxidation can produce rancid odours, colour changes and the loss of sensitive ingredients. No microorganism is involved, and a preservative is not designed to stop it.

Physical instability is a change in structure. FDA's shelf-life page lists several examples: emulsions, which are mixtures of water and oil, can separate; products can dry out, harden and crack; and temperature changes and exposure to sunlight and air can change colour, texture and smell. These are matters of emulsifiers, thickeners, processing and packaging, not of antimicrobial protection.

The three can interact. A separated emulsion can leave a water-rich layer in which preservatives are spread differently. Oxidation can produce compounds that change pH. A microbial problem can thin a product or create gas.

Preservative versus antioxidant versus stabilizer

Because the failures differ, so do the protective ingredients. A preservative targets microorganisms. An antioxidant, such as tocopherol (vitamin E) or another phenolic antioxidant, slows the oxidation of unsaturated oils that changes colour and odour. Chelating agents bind trace metal ions. That can slow some oxidation reactions and, as discussed later, can help some preservatives work. Stabilizers in the broad sense, meaning emulsifiers, thickeners and structuring agents, hold the physical form together.

Some ingredients do more than one of these jobs, which is a source of confusion. A plant extract with antioxidant activity is not automatically a preservative. When a label or a supplier says an ingredient "protects" the formula, the useful follow-up question is: protects against what, and how was that shown in this product?

Why appearance and smell cannot establish microbiological safety

It is tempting to think that if a cream looks and smells normal, it must be fine. The ICCR FAQs are direct on this point: an inadequately preserved product may change, but microorganisms in it may also remain unnoticed, and a consumer may not be able to tell (ICCR FAQs, PDF). FDA encourages consumers to report a bad smell or other signs of contamination even without a reaction. A sign can indicate a problem; its absence proves nothing.

For developers, the lesson is that sensory checks are useful quality observations, not microbiological evidence. A stability sample that still looks perfect after months in a warm cabinet says something about physical and some chemical stability. It says nothing on its own about whether the preservation system controls microorganisms introduced during use.

Preservation as a Whole-Product System

If preservation were only about the preservative, every cream with the same preservative would be equally protected. It is not.

Formula, manufacturing hygiene, packaging and expected use

A formulator choosing preservatives weighs the product's other ingredients, its intended performance, the pack, the part of the body it goes on and how people will use it. The general aim is the lowest level that works reliably.

It helps to treat the system as four layers:

1. Formula environment: Does the product itself make growth easy or hard? This covers water, pH and where the preservative sits.
2. Starting quality: How many and which microorganisms are present when the product is made?
3. Barrier: How easily can new microorganisms get in once the product is filled?
4. Use pattern: How often, by whom, where and with what is the product touched? Fingers, applicators, shower water and shared use all count.

A strong preservative system usually means that no single layer has to carry the entire load. A water-rich cream in a jar puts heavy demands on layer one, because layers three and four are weak. The same cream in a pump moves part of the burden to the pack. A dry powder used with a damp sponge has a strong first layer, which the use pattern can undo.

Why adding one material does not settle protection

Adding a stronger or larger amount of preservative is not a reliable way to fix a weak system, for several reasons. First, preservatives differ in which organisms they control and under what conditions. Second, other ingredients can interfere. Third, safety limits apply. Regulators set conditions on preservative use, and the Contact Dermatitis review cited above links preservative allergy to contact eczema and argues that some products may be over-preserved. Fourth, a preservative cannot clean up a heavily contaminated batch or protect a product that is repeatedly diluted with tap water.

This is why the effectiveness of a preservation system is checked during development and again on the finished product in its final pack.

Who is responsible for which part

In the United States, FDA states that companies and individuals who manufacture or distribute cosmetics are responsible for their products' safety, including making sure products are free from harmful microorganisms. The Modernization of Cosmetics Regulation Act of 2022 (MoCRA) adds a requirement that the responsible person maintain records supporting adequate safety substantiation (FDA, MoCRA). In the European Union, the safety of each product is supported by a cosmetic product safety report, and the Commission notes that any ingredient's use must be supported by a safety assessment of the product (European Commission, CosIng).

In a typical OEM or ODM project, the work splits roughly as follows; the exact split belongs in the project agreement.

The point is that "is it preserved?" is really five questions with five owners, and the weakest answer sets the limit.

pH, Water Activity and the Product Environment

Before any preservative is added, the formula itself sets the conditions microorganisms face. pH and water activity matter most, and neither works as a simple switch.

Water content versus water activity

Water content is how much water a product contains. Water activity, written a_w, is how much of that water is free for microorganisms to use. It is defined as the water vapour pressure over the product divided by that of pure water at the same temperature. The scale runs from 0 (completely dry) to 1 (pure water).

The two can differ a lot. Ingredients that bind water, such as glycerin and other humectants, sugars, salts and some polymers, lower water activity without removing any water. In a water-in-oil emulsion the water is held as droplets inside a continuous oil phase, which changes how available it is to microbes arriving at the surface.

The DGK microbiology group cautions that water activity in complex cosmetics should be measured, rather than inferred from water content or ingredient lists (DGK, 2025 technical paper).

Put together, this explains a pattern that runs through the rest of this guide. For an ordinary lotion or cream, water activity helps a little and preservation has to come mainly from elsewhere. For products that are dry by design, such as powders, sticks, balms and solid bars, low water activity can be the main protection, as long as the product stays dry.

Why pH, phase location and temperature interact

pH. Very acidic or very alkaline conditions inhibit many microorganisms, but most leave-on skin products are formulated in a comfortable, mildly acidic to near-neutral range. Products near neutral are generally more open to microbial growth. In that range, pH supports preservation but rarely carries it.

pH also changes how some preservatives work. A weak-acid preservative largely in its active form at pH 4 can be mostly inactive at pH 6, as the organic-acid table in the next chapter shows.

Phase location. In an emulsion, enough preservative must remain available in the aqueous phase where microbes can grow. Partitioning, micelles and adsorption can reduce that availability (Halla et al., 2018, Molecules).

Temperature. Warmth speeds up microbial growth and can break down some preservatives; FDA warns consumers that heat can make preservatives break down. Water activity readings also shift with temperature, which is why they are taken at a stated temperature.

Combining partial barriers is known as "hurdle technology". A review in the International Journal of Cosmetic Science applies it to cosmetics, naming manufacturing practice, packaging, emulsion type, low water activity and pH as combinable variables (Varvaresou et al., 2009, abstract).

No single universal safe threshold

An experimental study by Berthele and colleagues varied ethanol, pH and water activity together and checked the formulas with a challenge test to ISO 11930. Its abstract reports that, in the study's conditions, no single parameter at the tested values could guarantee microbiological stability. It also reports that some combinations stayed stable at a water activity above 0.6, even 0.8, without added preservatives (Berthele et al., 2014, abstract). These are specific experimental formulas. They show that the factors interact; they do not give a recipe.

ISO 29621 provides a low-risk assessment framework. DGK cautions that formula and use conditions belong in that assessment; no isolated water-activity value guarantees safety (DGK, 2025 technical paper).

Dry by design is not dry in use

Low-water and anhydrous products are protected by dryness only for as long as they stay dry. That protection is easy to lose in ordinary use:

Illustrative examples; no measurements implied.

In each case the formula as made may be low risk; the question is whether it stays so in use. Spores from dry raw materials can survive dryness and germinate once water is available. "Anhydrous" describes the formula on the day it was made, not the product on a bathroom shelf in month four.

Major Preservative Families and Their Roles

This chapter describes the preservative families readers are most likely to see on labels. It is a reference to roles and legal status, not a selection guide, and it gives no use levels for any product.

In the EU the word "preservative" has a legal meaning. Regulation (EC) No 1223/2009 defines this role by a substance's main intended purpose: inhibiting microbial development within the cosmetic. Only substances listed in Annex V may be used for that purpose, under the conditions written there. Where this chapter quotes an EU legal condition, it comes from the consolidated text of the Regulation dated 1 May 2026 (consolidated Regulation 1223/2009). (That text is a documentation tool; the binding versions are the acts in the Official Journal.) Annex V figures are legal ceilings in the ready-to-use product. They are not recommended levels and say nothing about whether a formula is adequately preserved.

Phenoxyethanol-based systems

Phenoxyethanol is a glycol ether used across skin-care, hair-care and make-up products, mainly active against bacteria and usually paired with other ingredients. Its current EU entry is Annex V entry 29, with a maximum of 1.0% in the ready-for-use preparation. That figure matches the conclusion of the Scientific Committee on Consumer Safety (SCCS) in opinion SCCS/1575/16 (final version October 2016). The committee "considers 2-phenoxyethanol safe for use as a preservative with a maximum concentration of 1.0%", having specifically examined children's exposure (SCCS opinion on phenoxyethanol). The opinion is scientific advice and the Annex V entry is the legal condition; here they coincide, but for butylparaben, below, they currently do not.

Phenoxyethanol can be paired with ethylhexylglycerin. A laboratory study by Langsrud and colleagues tested the combination; four co-authors were employees of preservative supplier Schülke & Mayr, which also participated in study design, interpretation and the publication decision. The researchers exposed E. coli to concentrations that were each sub-lethal alone and found rapid killing when the two were combined. In their experiments, adding ethylhexylglycerin to phenoxyethanol at 1:9 had a similar effect on membrane damage and viability as doubling the phenoxyethanol (Langsrud et al., 2016, PLoS One). That is a laboratory result on one bacterium; it explains a mechanism, not how a given cream will perform.

Paraben families

For a focused explanation of the four common names, illustrative ingredient labels and the limits of “paraben-free” claims, see parabens in cosmetics.

Parabens are esters of 4-hydroxybenzoic acid. FDA names methylparaben, propylparaben, butylparaben and ethylparaben as the most commonly used in cosmetics, and notes that labels typically list more than one paraben, often alongside other preservatives (FDA, Parabens in Cosmetics).

Because parabens attract the most debate, it helps to separate what the law says from what scientific committees have concluded.

• United States. FDA has no rules specific to preservatives in cosmetics; they are treated like any other cosmetic ingredient. On parabens it states: "At this time, we do not have information showing that parabens as they are used in cosmetics have an effect on human health."
• EU legal conditions (consolidated text, 1 May 2026). Annex V splits parabens into two entries. Entry 12 covers 4-hydroxybenzoic acid, methylparaben and ethylparaben and their salts: 0.4% (as acid) for a single ester and 0.8% (as acid) for mixtures of esters. Entry 12a covers butylparaben and propylparaben and their salts: 0.14% (as acid) for the sum of the two, within the 0.8% overall limit for mixtures. Entry 12a also says these may not be used in leave-on products designed for the nappy area of children under three. Isopropylparaben, isobutylparaben, phenylparaben, benzylparaben and pentylparaben appear in Annex II, the list of prohibited substances.
• EU scientific opinions. For methylparaben, SCCS opinion SCCS/1652/23 (December 2023, with a February 2024 corrigendum) concluded that use up to the existing entry 12 limits is safe (SCCS opinion on methylparaben). For butylparaben, opinion SCCS/1674/25 (final version 30 April 2025) concluded that 0.14% across all the product types it assessed, used in combination, was not safe for the child age groups evaluated. It found that lower concentrations in some product types would be safe (SCCS opinion on butylparaben, children's exposure). The consolidated text of 1 May 2026 still shows the entry 12a figure of 0.14%. An SCCS opinion feeds later rule-making but does not change the Annex by itself, and a conclusion on one ester cannot be transferred to another.

"Parabens" is therefore not one substance with one status: in the EU two common esters share one set of conditions, two share a tighter set and five are prohibited.

Benzoate, sorbate and other organic-acid systems

Organic acids and their salts, many familiar from food, appear on labels as Benzoic Acid, Sodium Benzoate, Sorbic Acid, Potassium Sorbate, Dehydroacetic Acid or Sodium Dehydroacetate. Their defining feature is pH dependence. These are weak acids, and the uncharged (undissociated) form is generally regarded as the one that crosses microbial membranes most readily. PubChem lists a dissociation constant (pKa) of about 4.2 for benzoic acid and about 4.76 for sorbic acid at 25 °C (PubChem, benzoic acid; PubChem, sorbic acid). From the standard Henderson–Hasselbalch relationship, the share of acid left in the undissociated form falls steeply as pH rises:

Editorial calculation from published pKa values, rounded; real formulas also involve buffering, solvents and partitioning.

The table explains a common development surprise. A formula that works well at pH 5 can lose most of its undissociated acid if it is adjusted to pH 6 for a milder feel, with no visible change at all.

The EU sets separate conditions for each. Sodium benzoate and benzoic acid (Annex V entry 1) are limited by product type: 2.5% (as acid) in rinse-off products other than oral products, 1.7% in oral products and 0.5% in leave-on products. Sorbic acid and its salts (entry 4) are limited to 0.6% (as acid). Dehydroacetic acid and its sodium salt (entry 13) are limited to 0.6% (as acid) and may not be used in aerosol sprays.

Benzyl alcohol-based systems

Benzyl alcohol (INCI: Benzyl Alcohol) is an aromatic alcohol with more than one legal identity, which is why it causes confusion on labels.

As a preservative, it was already listed in Annex V of Regulation 1223/2009 as adopted in 2009, as entry 34, with a 1.0% maximum. That remains the entry in the 2026 consolidated text (Regulation 1223/2009 as adopted, PDF). Benzyl alcohol is also listed in Annex III, the list of restricted substances, for uses other than inhibiting microorganisms, such as a solvent or fragrance component. That entry requires its presence to be named in the ingredient list above 0.001% in leave-on products and 0.01% in rinse-off products.

So the same INCI name can appear for different reasons, and the label alone does not show whether it belongs to the preservation system.

Isothiazolinones

Methylisothiazolinone (MI) and the mixture methylchloroisothiazolinone/methylisothiazolinone (MCI/MI) are effective at very low concentrations. They are also strongly associated with contact allergy, and EU conditions reflect that.

In SCCS opinion SCCS/1557/15 (June 2015), the committee concluded that the data did not support MI's safe use at 100 ppm in rinse-off or leave-on hair products from the point of view of inducing contact allergy. It considered 15 ppm (0.0015%) safe for rinse-off products on that basis (SCCS opinion on MI). The consolidated EU text now limits both MI (Annex V entry 57) and the MCI/MI mixture (entry 39) to rinse-off products only, at a maximum of 0.0015%. In the EU, neither may be used as a preservative in leave-on cosmetics.

Formaldehyde releasers, current and removed

Formaldehyde releasers are substances that slowly give off small amounts of formaldehyde, the actual antimicrobial agent. A peer-reviewed review cites research showing that the amount released depends on the product matrix, pH, storage time and especially temperature (Halla et al., 2018, Molecules).

The EU now separates this family into substances still listed and substances that have been removed:

• Still listed in Annex V (consolidated text, 1 May 2026): among others, DMDM Hydantoin (entry 33, 0.6%), Imidazolidinyl Urea (entry 27, 0.6%) and Diazolidinyl Urea (entry 46, 0.5%). The Annex V preamble adds a labelling rule. Finished products containing formaldehyde-releasing substances must carry the warning "releases formaldehyde" where total released formaldehyde exceeds 0.001% (10 ppm).
• Removed and prohibited: Commission Regulation (EU) 2019/831 deleted Quaternium-15 from Annex V and added it to Annex II, the prohibited list. Its recitals explain why. The cis-isomer in Quaternium-15 had been classified as a category 2 CMR substance, and the SCCS had concluded that its safety in cosmetics could not be established. The same regulation removed formaldehyde and paraformaldehyde from Annex V and added them to Annex II (Regulation (EU) 2019/831, Official Journal text).

Quaternium-15 is therefore a historical example in the EU, not a current option. For the releasers still in use, EU claims guidance adds a labelling point. A "free from formaldehyde" claim is not acceptable if the product contains a formaldehyde-releasing ingredient such as Diazolidinyl Urea (Technical document on cosmetic claims, 2017, PDF).

Reading this chapter as a whole

The EU column is not a global table; other markets keep their own lists.

Why Blends Behave Differently From Single Ingredients

Most finished cosmetics do not rely on a single preservative. The label may show two, three or more antimicrobial ingredients, sometimes alongside materials not usually thought of as preservatives at all. This chapter explains why.

Spectrum, phase distribution and solubility

One reason is spectrum. Some preservatives are stronger against bacteria, others against yeasts and moulds. Combining them covers more ground, which is why FDA notes that parabens are often used with other preservatives.

Another reason is mechanism. If two ingredients damage microorganisms in different ways, the combination can work at levels where neither would work alone. Laboratory work on ethylhexylglycerin and phenoxyethanol, described in the previous chapter, shows one such case. Ethylhexylglycerin weakened the bacterial membrane, and phenoxyethanol then killed cells at a concentration that was sub-lethal on its own.

Multifunctional ingredients and boosters

Some ingredients are added mainly for another job but also contribute antimicrobial activity. They are not listed as preservatives in the EU sense, yet they can carry part of the load.

• 1,2-alkanediols, such as pentylene glycol, 1,2-hexanediol and caprylyl glycol, act through their amphiphilic structure, which lets them insert into microbial membranes. A study by Sigg and Daniels tested four alkanediols in a standard non-ionic cream using the European Pharmacopoeia preservation test. Antimicrobial activity rose with the length of the carbon chain. So did interference with the cream's structure: longer-chain alkanediols integrated into the cream's lamellar network and slightly reduced its consistency and stability (Sigg and Daniels, 2020, Pharmaceutics). The trade-off is typical of this group. The property that makes an ingredient antimicrobial can also change the texture.
• Chelating agents, such as disodium EDTA, tetrasodium glutamate diacetate (GLDA) or phytic acid, bind metal ions. The 2018 Molecules review describes how they can make bacterial outer membranes more permeable to other antimicrobials and limit the iron microbes need.
• Medium-chain glyceryl esters, such as glyceryl caprylate, are emulsifiers or emollients with reported activity against some bacteria and fungi.
• Humectants such as glycerin, at high enough levels, make a formula harder to spoil, partly by lowering water activity.

Ethylhexylglycerin belongs in this "booster" group. These materials are often what makes a "preservative-free" label possible, which the next chapter discusses. They also make label reading harder, because their antimicrobial role is not declared anywhere on the pack.

Supplier blend data versus finished-product performance

Ingredient suppliers publish data on their preservative blends, often including spectrum testing, compatibility notes and suggested product types. This information is valuable for choosing what to test. It is not evidence that a specific finished product is adequately preserved.

The gap exists because the supplier tested the blend under its own conditions, in its own model formulas, sometimes against a limited set of organisms. Your product has different surfactants, oils, actives, fragrance, pH and pack, and any of them can strengthen or weaken the blend's effect. The alkanediol study above shows how large that gap can be in the other direction too: an ingredient chosen for protection changed the structure of the cream it was added to. Both effects only show up in the actual formula. Supplier claims such as "pH independent" or "effective at low dosage" describe a material, not your product.

Natural, Naturally Derived and Preservative-Free Claims

People search for "natural preservatives for cosmetics" because they want gentler or simpler products. Confusion starts when words about origin or marketing are read as statements about safety or protection.

What origin terms do and do not establish

"Natural", "naturally derived", "nature-identical" and "organic" describe where an ingredient comes from or how it was produced. They say nothing direct about whether a product is protected from microorganisms.

Regulators are consistent on this. FDA's page on organic cosmetics states that an ingredient's source does not determine its safety. It notes that many plants, organically grown or not, contain substances that may be toxic or allergenic, and that all cosmetics must meet the same safety requirement whatever their source (FDA, Organic Cosmetics). The ICCR FAQs make the same point for preservation: natural and organic products need the same level of preservation testing as any other.

In the United States, FDA notes that "organic" is not defined in the laws it enforces for cosmetics. The term is governed separately by the US Department of Agriculture's National Organic Program for agricultural ingredients. Certification schemes can narrow which preservatives a brand may use, but certification is not a preservation result. Our guide to organic and natural certifications for cosmetics covers those schemes and why preservation is often the real constraint in organic formulas.

Tocopherol, essential oils and botanical extracts

Three groups of plant-associated ingredients are often presented as "natural preservatives". Their roles differ.

Tocopherol (vitamin E) is an antioxidant. It helps slow the oxidation of oils, which is the second failure type described earlier. It does not control bacteria or fungi. A formula with tocopherol and no antimicrobial protection may be well protected against rancidity and poorly protected against microbes.

Essential oils can show antimicrobial activity in laboratory tests, but that does not establish adequate preservation of a finished cosmetic (Halla et al., 2018, Molecules). Separate EU fragrance-allergen rules also apply to relevant constituents; check the specific material and product.

Botanical extracts. Ask what preparation the supplier provides, including carriers or added preservatives, and what evidence relates to that exact material.

Multifunctional systems and "preservative-free" wording

Many products labelled "preservative-free" or "self-preserving" are not unprotected. They are usually built on the multifunctional approach described in the previous chapter: a combination of low water activity, suitable pH, alkanediols, glyceryl esters, chelators and protective packaging, none of which appears in the legal list of preservatives. The 2009 review on self-preserving cosmetics cited earlier describes this hurdle-based design.

The EU has specific guidance on these claims. The Technical document on cosmetic claims, agreed by an EU working group of Member States and stakeholders, is not legally binding, as it says itself, but national authorities use it as a reference. It sets out several points about "free from" claims (Technical document on cosmetic claims, 2017, PDF):

• A product claiming to be free from a functional group defined in the Regulation, such as preservatives, should not contain any ingredient that belongs to that group.
• "Free from" claims about functional groups should generally not be allowed if the product contains multifunctional ingredients that also perform that function. The document notes possible exceptions, for example based on challenge test results for the formula without those ingredients.
• It gives the example of fine fragrances, which usually contain enough alcohol that added preservatives are unnecessary, and says highlighting "no preservatives" there would be dishonest.
• It describes "low in allergens because without preservatives" as an unfair claim, because it assumes all preservatives are allergenic.

Other markets have their own rules. The working-group document shows the direction of regulatory thinking: "preservative-free" is an evidence-backed claim about formula composition, not a shortcut to "gentle" or "safe".

What this means for people searching for natural preservatives

For a consumer, a helpful way to read a "natural" or "preservative-free" product is to look at the format and storage advice. A water-rich cream in a jar with no recognisable preservation system and a long PAO deserves a question. A dry balm, a single-use sachet or a high-alcohol fragrance may have good reasons to need little or no added preservative.

For a brand, the path to a more natural or minimal system is a development project, not a substitution. It typically means starting from the product format and pack, choosing materials whose role is documented, and accepting that evidence for the finished product is needed. This guide does not suggest specific natural preservatives or doses; improvised home-made preservation is where the risks in this guide are greatest. FDA links a Small Business & Homemade Cosmetics Fact Sheet from its microbiological safety page, a reasonable starting point for anyone in the United States making products to sell.

Packaging and Use Change the Protection Question

Two identical batches of the same cream can face very different challenges depending on what they are filled into: an open jar invites fingers into the product every day, while a pump dispenses without the user touching the bulk.

Jars, tubes, pumps, airless systems and single-use formats

One frequently cited early study tested this directly. Brannan and Dille packed unpreserved shampoo and skin lotion aseptically into containers with different closures and gave them to volunteers for normal use over two to three weeks. In the abstract, the lotion was contaminated in 71% of screw-cap containers, 39% of flip-cap containers and 10% of pump-top containers. For shampoo, flip-cap closures gave the best protection, with no contamination recorded (Brannan and Dille, 1990, abstract). The study used deliberately unpreserved products, small groups and specific packs from that time, so its percentages should not be applied to modern products. Its direction is still clear: the closure shapes how much contamination gets in.

Assess each pack's opening, whether a used applicator returns to the bulk, and possible air or water backflow. Format names alone do not determine protection.

Applicator contact, repeated opening and refills

The way a product is applied adds its own route. Mascara wands, lip glosses with doe-foot applicators and dropper serums all return an applicator that has touched skin to the product. FDA notes that applicators, especially mascara wands, are exposed to bacteria and fungi every time they are used. It says manufacturers usually recommend discarding mascara two to four months after purchase. A dropper that touches the face and goes back into the bottle works the same way. Our packaging compatibility guide covers droppers and return contamination in more detail.

Where the product is used matters too. Products opened in the shower meet water and steam. Products shared at counters or in salons meet more people's skin.

Refills raise their own questions. A refill pod clicked into a permanent outer case may keep the product sealed. A pouch emptied into a reused bottle exposes it to whatever is left in that bottle, plus air and handling during the transfer. A refill system needs its own assessment of transfer, cleaning of reused parts and time in use; it cannot simply inherit the original pack's evidence.

A closed pack is one design factor, not a safety result

Some container materials can absorb oil-soluble preservatives, reducing the amount available in the product; this is one reason to assess compatibility in the actual pack (Halla et al., 2018, Molecules).

So "airless" on a brief is a useful design choice. It is not a substitute for evidence about the product in that pack. If packaging is part of the argument for a lighter preservation system, the evidence should be produced in the actual pack and should look at how the pack behaves in use. Our quality and testing page explains why packaging evidence has to attach to the final formula and the complete component assembly, not to a generic material name.

Manufacturing Hygiene and Raw-Material Quality

Control before the product reaches a consumer

In the EU, Article 8 of Regulation 1223/2009 requires cosmetics to be manufactured in line with good manufacturing practice, and compliance is presumed where manufacture follows the relevant harmonised standard; in practice that standard is ISO 22716 (consolidated Regulation 1223/2009). Cosmetic production is not sterile, so even a well-run plant releases products with some low level of microorganisms. Hygiene keeps that starting level low and predictable. Preservation then has to keep it from growing and deal with what users add. Neither replaces the other.

In practice, hygiene covers incoming material checks, water system control, cleaning and sanitising of equipment between batches, controlled hold times for bulk product, clean filling, and staff practices. In the United States, FDA has published draft guidance on cosmetic GMP, informed by ISO 22716. FDA marks it "Not for implementation", notes that it is non-binding, and says it intends to withdraw or revise it based on the GMP rulemaking required by MoCRA (FDA, Draft Guidance on Cosmetic GMP). It is background, not final US rules.

What preservatives cannot compensate for

A preservative system is designed to deal with ordinary low-level contamination and with what users introduce. It is not designed to sterilise a batch made with contaminated water or left uncovered in a warm tank. It cannot reliably deal with a raw material carrying a heavy load of resistant organisms. And it cannot neutralise the biofilm that forms in poorly cleaned pipework and seeds batch after batch.

A system shown to work on laboratory samples may also meet a different challenge at full scale, where larger equipment, longer hold times and different filling lines change the starting picture.

Questions a brand can reasonably ask

A brand does not need to audit a factory's hygiene programme line by line to ask useful questions. It can ask how incoming raw materials with higher microbial risk, such as botanical extracts, powders and gums, are checked before use. It can ask how the process water is controlled and monitored, how long bulk product may be held before filling and under what conditions, and whether batch microbiological results are available for each released batch.

Where the detail lives

Our cosmetic manufacturing process guide walks through raw-material release, bulk hold, filling and batch release step by step, including the records a brand can ask to see.

What Different Tests Can and Cannot Show

Several kinds of evidence are relevant to preservation. Each answers a different question, and confusing them is one of the most common mistakes in product files.

Batch microbiology, challenge assessment and stability at a high level

Batch microbiological quality testing asks how many and which microorganisms are present in a given batch at the time it is tested. It is a release check. It says nothing about how the product will cope with contamination introduced later. FDA publishes microbiological methods for cosmetics as Chapter 23 of its Bacteriological Analytical Manual (FDA BAM Chapter 23).

Preservative efficacy (challenge) evaluation asks whether a preservation system can control microorganisms deliberately introduced under defined laboratory conditions. ISO 11930 excludes products assessed as microbiologically low risk under its Annex A and ISO 29621 (ISO 11930 public scope). The need for challenge evaluation and routine microbial testing should follow a documented product-specific risk assessment, applicable market requirements and the quality programme. Neither test should be presented as automatically required for every low-risk product.

Stability evaluation asks how the product changes over time, often under raised temperatures and in its intended pack. It deals mainly with the physical and chemical failures described earlier, and supports decisions on shelf life. Our stability testing guide explains how to read those reports.

None of these is a "safety test" that approves a product. FDA states that neither the law nor FDA regulations require specific tests to demonstrate the safety of individual products or ingredients, while the responsible person must keep records supporting adequate safety substantiation. In the EU, the evidence feeds the product safety report. Our cosmetic safety testing guide explains how the pieces fit together.

Sample and version linkage

A result describes the sample that was tested. If the formula, a key raw-material grade, the pH, the fragrance or the pack changes afterwards, the result may no longer apply. That does not mean every change needs a new study. It does mean someone qualified should review whether the earlier evidence still applies and record the decision. Good records of which formula version, batch and pack were tested make that review possible.

When to compare study types

Laboratories offer challenge work in very different forms. A quick development screen and a full study to a named method can both be called a "challenge test" while delivering different things. Our separate guide, Preservative Challenge Testing: Screening vs a Full Study, explains how to compare proposals, what a usable report contains, and why the recovery question matters. This guide does not repeat that comparison.

Shelf Life, PAO and the Final Pack

Labels in many markets give two different kinds of time information. They are easy to confuse.

Unopened durability versus after-opening use period

Shelf life, in FDA's usage, is broad. FDA describes it as the length of time you can expect a product to look and act as expected and to stay safe for use, and says it varies with product type, use and storage (FDA, Shelf Life and Expiration Dating).

Date of minimum durability is a precise EU legal term. Article 19(1)(c) of Regulation 1223/2009 defines it as the date until which the product, stored under appropriate conditions, will continue to fulfil its initial function and remain safe. When it is used, it appears on the label after an hourglass symbol or the words "best used before the end of".

Period after opening (PAO) is the time after opening during which the product is safe and can be used without harm. On EU labels it is the open-jar symbol followed by a number of months or years, such as "12M".

How the EU 30-month rule works, including its exception

Article 19(1)(c) links these terms in one rule (consolidated Regulation 1223/2009):

1. A product whose minimum durability is 30 months or less must show its date of minimum durability.
2. A product whose minimum durability is more than 30 months does not have to show that date. Instead it must show the PAO.
3. The PAO is shown except where the concept of durability after opening is not relevant.

The exception does real work. Recital 48 of the Regulation gives the intended scope: the PAO requirement should not apply to single-use products, products not at risk of deterioration, or products which do not open (Regulation 1223/2009 as adopted, PDF). Some formats that usually fall within it, judged product by product:

• Single-use products, such as a sealed one-application sachet or ampoule. Once opened, it is used up, so there is no "after opening" period to manage.
• Products that do not open in the ordinary sense. Many pressurised aerosols are an example: the user dispenses product but never exposes the bulk to air or fingers.
• Products not at risk of deterioration after opening, where the safety assessment shows that opening does not start any meaningful change.

The exception is a reasoned conclusion about a specific product, not a format shortcut. An airless pump, for instance, is not automatically a product "which does not open", since product still passes through a valve. Whether the concept is relevant is decided in that product's safety assessment, and the reasoning belongs in the file.

Ireland's Health Products Regulatory Authority (HPRA) adds a practical point. Durability dates and PAO "are not random figures": they are determined as part of the testing detailed in the cosmetic product safety report, and supporting evidence must be included in it (HPRA, cosmetic product labelling).

In the US, FDA states that cosmetics do not have a universal legal shelf-life or expiry-label requirement, though establishing shelf life remains the manufacturer's responsibility. Products regulated as drugs, including SPF moisturisers, follow drug stability and dating rules. FDA's sunscreen guidance explains an exception to expiry-date labelling when stability testing demonstrates at least three years of stability. Our labelling comparison sets out dating conventions market by market.

Why 6M, 12M and 24M are evidence questions

A PAO is sometimes treated as a label choice, picked to match a competitor. As HPRA's note makes clear, it is a conclusion the safety report has to support.

A defensible PAO draws on the whole system described in this guide. It considers how quickly the formula's sensitive ingredients degrade once air gets in, how well the preservation system copes with repeated contamination, and how much the pack protects. It also considers how long a typical user takes to finish the product. A 200 ml body lotion used daily and a 15 ml eye cream used sparingly have very different real-world open times, even if their formulas look alike. This guide gives no standard months for product types. The right figure depends on the evidence for each product.

Packaging or formula changes and relevance review

Because shelf life and PAO rest on evidence about a specific formula in a specific pack, changes can affect them. Moving from a jar to an airless pump might support a longer PAO. Moving the other way, or changing the preservative system, fragrance or pH, might shorten it. In each case the question is the one raised in the previous chapter: does the existing evidence still apply, and who has reviewed that? Changing the number on the label without that review undermines the whole basis of the figure.

FDA's advice to consumers adds the user's side. Store products as directed. Throw away mascara that has dried rather than adding water or saliva. Be careful with products from flea markets or resellers, which FDA says may be past their shelf life, used, diluted, tampered with or counterfeit.

Three Illustrative Protection Scenarios

The scenarios below are illustrative decision maps, not case studies; no real product, customer, measurement or test result is described. They show how the factors in this guide combine.

Scenario A: a water-rich cream in a jar versus a pump

A brand wants a rich face cream, mostly water with plant oils and a humectant, and likes the look of a wide glass jar. It is also considering an airless pump.

The comparison does not say the jar is wrong. Many good creams are sold in jars; it shows that the jar puts more of the load on the formula. Switching to the pump later is a change that needs its own evidence review. A spatula, a smaller jar size or a shorter PAO are other ways a team might reduce the challenge.

For a cream project, our moisturizer manufacturing page is the commercial starting point for discussing texture, packaging and development requirements.

Scenario B: a rinse-off cleanser with pH and fragrance changes

A gel cleanser is already developed, with a preservation system built around an organic acid salt at mildly acidic pH. Late in development, the brand asks for two changes: a higher pH to make the cleanser feel milder, and a new fragrance.

Each change touches the preservation system. The organic-acid table in Chapter 6 shows the scale: moving a sorbate system from pH 5.0 to 6.0 would cut the undissociated share from roughly a third to about a twentieth, with no visible sign. A new fragrance brings new ingredients, some of which may interact with the preservative or with the surfactants. The bottle will also be opened in the shower, where water can get in.

The sensible sequence is to treat the revised formula as a new version. The formulator reviews whether the existing preservation system still fits the new pH or needs adjustment. The team decides with the laboratory or assessor what evidence the change requires, and records it. What the team should not do is assume the earlier result covers the new version because "only the pH and scent changed".

To develop a rinse-off facial cleanser, see our cleanser manufacturing page and include the proposed pH, fragrance and dispenser in your brief.

Scenario C: a low-water or refill concept

A brand is developing a solid face-cleansing bar with no added water, plus a refillable aluminium case. Marketing would like to say "preservative-free".

Low water activity may well be the main protection here. The "dry by design" table earlier in this guide shows how that protection can be lost. A bar left wet in a case that does not drain spends hours at a time with a damp surface. The refill case adds questions about how it is cleaned and dried between bars.

In all three, the protection strategy is a set of decisions with evidence attached, not a property of one ingredient or one pack. Working through format, pack and evidence together from the first brief is part of how an ODM development project is scoped.

Ingredient Labels and Market Boundaries

How to recognise the declared materials

Ingredient lists are the main window into a product's preservation system, with limits. In the EU, HPRA's summary states that all substances must appear in descending order of quantity, using INCI names. In the United States, FDA requires ingredients on retail cosmetics to be listed by their common or usual names, generally in descending order of predominance. FDA does not accept "Aqua" in place of "Water" except in parentheses after the English name (FDA, Cosmetic Ingredient Names).

Preservatives are usually used at low levels, so they tend to appear towards the end of the list. Their position tells you little about how much is present, and nothing about how well the system works. Multifunctional ingredients such as caprylyl glycol, glyceryl caprylate or chelating agents appear under their own names with no "preservative" label. Benzyl alcohol may be there as a solvent or fragrance ingredient. A trade-name blend appears as its individual INCI components.

US safety responsibility versus EU positive-list conditions

In the United States, FDA states that it has no special rules for preservatives in cosmetics. Cosmetics other than color additives do not need FDA approval before sale, but they must not be adulterated or misbranded. The company is responsible for safety.

In the European Union, preservatives are governed by a positive list. A substance may be used as a preservative only if it appears in Annex V of Regulation (EC) No 1223/2009, under the conditions written there. Some substances are prohibited outright in Annex II. The Commission's CosIng database is convenient for look-ups but, in the Commission's words, has "informative purpose and no legal value"; for preservatives, only substances authorised in Annex V are listed in CosIng as preservatives (European Commission, CosIng). Only the Regulation and its Annexes decide whether, and under which conditions, a substance may be used.

Other markets, including the ASEAN countries, China, Japan, Canada and others, have their own lists and conditions. Our guide to banned cosmetic ingredients in the EU explains how the EU annexes work together.

No globally universal maximum-concentration table

The EU figures in this guide come from the consolidated text of Regulation 1223/2009 dated 1 May 2026. The SCCS opinions are cited separately as scientific advice. Even within the EU, the legal entry and the newest scientific opinion can differ for a while, as the butylparaben example shows. Other markets publish their own lists, and the same substance can carry different limits, product-type restrictions or warnings in each. A product sold in several markets has to meet every one of them, and the strictest applicable rule often shapes the formula.

Common Questions and an Evidence Worksheet

The earlier chapters answer the most common questions: whether a normal-looking product can still be contaminated (Chapter 3), what "natural" and "preservative-free" mean (Chapter 8), and whether one passed test covers later versions (Chapter 11). The questions below are ones readers often ask that the main text has not yet addressed directly.

Are preservatives bad for skin?

Some preservatives can cause allergic reactions in people who have become sensitised to them. That is the main reason several, notably methylisothiazolinone, now face tight EU limits. Both too little and too much preservation carry risks. Too little exposes users to contaminated product; too much increases exposure to the preservative. Regulators therefore set conditions rather than simply banning or allowing whole families. A person with a suspected reaction should speak to a healthcare professional, and in the United States can report it to FDA.

Is a short PAO, such as 3 months, a sign of poor quality?

No single PAO indicates good or poor quality. It reflects product-specific evidence for the complete formula, packaging, expected use and storage. Counting listed preservatives, or seeing a multifunctional ingredient on an INCI list, cannot tell a reader which PAO is justified.

Is a product kept in the fridge better preserved?

Refrigeration does not establish microbiological safety or justify extending a product's expiry date or PAO. Follow the manufacturer's storage instructions, which should reflect the conditions assessed for that product. A fridge cannot replace adequate preservation or make a suspected contaminated product suitable for use.

Can I extend a product's life by decanting it into a cleaner container?

Usually not. Decanting exposes the product to air, hands and a container that has not been assessed with that formula. Any shelf-life or PAO evidence was produced for the original pack. Unless the manufacturer says otherwise, it is better to use the product in the pack it came in.

An evidence worksheet

The worksheet below connects the threads of this guide. It is a set of questions, not a checklist of required tests.

A Product-Development Brief and Next Steps

Preservation belongs in the first brief, not the last week before launch. A short self-check helps:

• Is the product water-rich, low-water or anhydrous, and does it stay that way in use?
• Will people touch the bulk product with fingers or an applicator?
• Will it be used in the shower or shared?
• Do any claims, such as "preservative-free" or a natural certification, limit which preservatives can be used?
• Which markets will it be sold in, and whose rules are strictest for the materials you prefer?

A useful brief states the product format, the intended pack and applicator, the target markets, any ingredient or claim constraints such as "paraben-free" or a natural certification, and the intended shelf life and PAO. It also asks the manufacturer two questions: what protection strategy they propose, and what evidence they will provide or arrange.

Our quality and testing page explains which checks we document and how external laboratories fit when a project or market needs them. Our ODM development service describes how a formula moves from brief to sample. When you know your product, pack and markets, you can request a quote. The method, laboratory and any external testing are agreed per project in writing.