Phenoxyethanol Versus Organic-Acid Preservative Systems

Cosmetics that contain water need protection against microbial contamination.¹ Two approaches that often appear on ingredient lists are PhenoxyethanolFunctionalPhenoxyethanolFunctionalPhenoxyethanol · PreservativeBroad-spectrum cosmetic preservative commonly capped around 1% in leave-ons.Open in the Index →, an aromatic ether,² and organic acids such as Benzoic AcidFunctionalBenzoic AcidFunctionalBenzoic Acid · Preservative / boosterBenzoic Acid is a preservative or preservative booster used to protect water-based formulas.Open in the Index →, Sodium BenzoateFunctionalSodium BenzoateFunctionalSodium Benzoate · Preservative / boosterSodium Benzoate is a preservative or preservative booster used to protect water-based formulas.Open in the Index →, Sorbic AcidFunctionalSorbic AcidFunctionalSorbic Acid · Preservative / boosterSorbic Acid is a preservative or preservative booster used to protect water-based formulas.Open in the Index → and Potassium SorbateFunctionalPotassium SorbateFunctionalPotassium Sorbate · Preservative / boosterPotassium Sorbate is a preservative or preservative booster used to protect water-based formulas.Open in the Index →. This map looks at the choice as a question of chemistry and evidence: how each system works, what regulators and safety panels have concluded, and what has to happen in the finished formula for either one to do its job.

This map covers three questions in turn: safety, legal limits and how well each works. It grades each claim on The Skindex scale of Strong, Moderate, Limited and Emerging. It is not a ranking, and neither system comes out as the universal answer.

Why preservatives are not optional

A 2018 review of cosmetic preservation notes that any product containing water and organic or inorganic compounds needs protection against microbial contamination, both for consumer safety and for shelf life.¹ The same review names Pseudomonas aeruginosa, Staphylococcus aureus and Candida albicans as the main potential pathogens in cosmetic products.¹

It also sets out a principle central to this comparison: a preservation system is validated in the finished product, through good manufacturing practice, raw-material control and a challenge test. The review describes the challenge test as the method adopted by international regulations: the product is deliberately exposed to test organisms, typically bacteria, a mould and a yeast, and the preservative effect is measured.¹ The name on the ingredient list does not tell you whether that test was passed. The test result does.

The safety lane: phenoxyethanol

At up to 1% as a preservative: Strong. Two independent reviews reach the same place. The US Cosmetic Ingredient Review (CIR) panel concluded in 1990 that phenoxyethanol is safe as a cosmetic ingredient in the then-current practices of use and concentration, noting use as a preservative below 1%. In clinical studies it was neither a primary irritant nor a sensitiser, and it was not phototoxic. The panel also recorded that the undiluted material is a strong eye irritant.²

In Europe, the Scientific Committee on Consumer Safety (SCCS) re-examined phenoxyethanol after the French medicines and health products agency, ANSM, submitted a risk assessment in 2012. ANSM had proposed lowering the limit to 0.4% in products for children under three and excluding it from nappy-area products.³ In its final opinion, adopted on 6 October 2016, the SCCS concluded that phenoxyethanol is "safe for use as a preservative with a maximum concentration of 1.0%."³ The committee described contact sensitisation in humans as documented but rare, with a very low risk of becoming sensitised.³

For children under three: Moderate. The SCCS agreed that young children may be more exposed than adults. Using reasonable worst-case scenarios, it calculated margins of safety of about 41 to 53, depending on the body weight assumed. It judged these sufficient after reducing one of its default adjustment factors on toxicokinetic grounds, which lowers the minimum acceptable margin from 100 to 25.³ That is a considered expert judgement resting on modelling choices, and the opinion states that it does not account for exposure from sources other than cosmetics.³ The grade reflects both the depth of the review and the fact that it turned on those assumptions.

The safety lane: organic acids

Benzoic acid and sodium benzoate: Moderate. The CIR's 2001 report, which also covered Benzyl AlcoholFunctionalBenzyl AlcoholFunctionalBenzyl Alcohol · Preservative / boosterBenzyl Alcohol is a preservative or preservative booster used to protect water-based formulas.Open in the Index →, concluded that these ingredients can be used safely at concentrations up to 5%. Benzoic acid was not a sensitiser at 2%. The report noted that these ingredients can produce nonimmunologic contact urticaria and immediate contact reactions, with wheals, redness and itching, and advised manufacturers to consider it when formulating for infants and children. The panel also judged the available data insufficient for products where inhalation is a route of exposure.⁴ The assessment dates from 2001.

Sorbic acid and potassium sorbate: Moderate. The CIR's 1988 report concluded both are safe as cosmetic ingredients in the practices of use and concentration then reviewed, with use of sorbic acid as a preservative reported at up to 1.0%. Formulations with up to 0.5% were not significant irritants or sensitisers in testing.⁵ The underlying data are older still.

Dehydroacetic acid: Moderate. Dehydroacetic Acid (DHA) and its sodium salt are also listed as preservatives in EU Annex V.⁶ The CIR first found them safe as used in 1985, reaffirmed that conclusion after a 2003 re-review, and reaffirmed it again in a review published in 2024. That latest review noted that reported uses had increased significantly since the previous re-review, and considered cases of hypersensitivity after use of creams containing sodium dehydroacetate. The panel said concern was mitigated by the minimal number of case reports compared with the widespread use of these ingredients, and by low use concentrations.⁷

The legal lane: what the EU allows

EU cosmetics law lists permitted preservatives in Annex V of Regulation (EC) No 1223/2009, each with a ceiling. The figures below are from the consolidated text dated 18 May 2026.⁶

• Phenoxyethanol: up to 1.0%.
• Benzoic acid and sodium benzoate: up to 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, including potassium sorbate: up to 0.6% (as acid).
• Dehydroacetic acid and sodium dehydroacetate: up to 0.6% (as acid), and not in aerosol sprays.
• Benzyl alcohol: up to 1.0% as a preservative. When it is used for other purposes, Annex III requires it to be named in the ingredient list above 0.001% in leave-on products and 0.01% in rinse-off products.

These are ceilings, not recommended doses. The organic-acid limits are expressed as the acid.⁶

The efficacy lane: why formula design decides the outcome

Organic acids depend on pH: Moderate. The 2018 review identifies pH as a major determinant of organic-acid effectiveness, because pH controls how much of the acid is present in its uncharged form, the form that can enter microbial cells. It describes several mechanisms, including acidifying the cell interior, and warns that pH changes can cause a progressive loss of antimicrobial activity.¹ In practical terms, the pH of the finished formula matters for how well an organic acid works. The review is a synthesis rather than a set of head-to-head trials, which is why the grade stops at Moderate.

Phenoxyethanol depends on where it ends up: Limited. A 2018 laboratory study in oil-in-water emulsion gels found that phenoxyethanol's antimicrobial effect tracked its free concentration in the water phase. The more lipophilic the oil phase, and the lower the content of dissolved ingredients in the water phase, the more phenoxyethanol went to the water phase and the better it worked. Higher emulsifier levels reduced the free amount through binding in micelles. Aspergillus brasiliensis, a mould, was the most resistant organism tested and Staphylococcus aureus the most sensitive.⁸ That is one study in one family of formulas, but it makes a useful point: the label percentage is not the same as the amount available to act.

Boosters can change the arithmetic: Emerging. In a 2016 laboratory study, EthylhexylglycerinFunctionalEthylhexylglycerinFunctionalEthylhexylglycerin · Preservative booster / skin-conditioningAlkyl glyceryl ether used to boost preservatives and as a mild skin-conditioning agent.Open in the Index → (EHG) at a concentration that did not kill bacteria on its own damaged cell membranes and made phenoxyethanol far more lethal to E. coli. The authors reported that adding EHG to phenoxyethanol in a 1:9 ratio had a similar effect on membrane damage and bacterial survival as doubling the phenoxyethanol concentration.⁹ The study was in-vitro work with a single organism, and four of its six authors were employees of Schülke & Mayr GmbH, as the paper's competing-interests statement declares.⁹

Head-to-head: not graded. None of the sources reviewed here compared a phenoxyethanol system with an organic-acid system in the same finished product. Without that, any claim that one family is simply more effective than the other goes beyond the evidence.

FAQ

Is phenoxyethanol unsafe for babies?

The SCCS reviewed exactly this concern, raised by the French authority ANSM, and concluded in 2016 that phenoxyethanol is safe as a preservative at up to 1.0%, with the margins calculated for children under three judged sufficient.³ The opinion does not cover exposure from non-cosmetic sources.

Are organic-acid preservatives gentler?

Not automatically. The CIR found benzoic acid non-sensitising at 2% but noted it can trigger nonimmunologic contact urticaria.⁴ Phenoxyethanol sensitisation is documented but rare, according to the SCCS.³ Tolerance depends on the whole formula.

Why do some products use several preservatives at once?

Each ingredient has limits, whether pH dependence for organic acids or partitioning between oil and water for phenoxyethanol.⁸¹ The authors of the EHG study describe potentiating agents as a way to keep preservative concentrations as low as possible, and the challenge test shows whether the blend works.⁹¹

Can I tell from the label whether a product is well preserved?

1. The label shows which preservatives are present, not whether the formula passed a challenge test at its actual pH and oil content.¹

Where the evidence stops

Phenoxyethanol at up to 1% rests on a detailed 2016 SCCS opinion as well as a CIR review. The benzoates and sorbates rest on older CIR conclusions, with EU ceilings that for benzoic acid are lower in leave-on products, and DHA was reaffirmed as safe in a review published in 2024.³²⁴⁵⁷⁶ On efficacy, the evidence shows that both approaches depend on formula design, pH for the acids and free water-phase concentration for phenoxyethanol, and that neither has been shown to beat the other in a like-for-like product.⁸¹

Footnotes

1. Halla N, Fernandes IP, Heleno SA, et al. Cosmetics Preservation: A Review on Present Strategies. Molecules. 2018. PMID 29958439. doi:10.3390/molecules23071571 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

2. Cosmetic Ingredient Review. Final Report on the Safety Assessment of Phenoxyethanol. J Am Coll Toxicol. 1990. doi:10.3109/10915819009078737 ↩ ↩² ↩³

3. Scientific Committee on Consumer Safety. Opinion on Phenoxyethanol, SCCS/1575/16, final version adopted 6 October 2016. https://health.ec.europa.eu/publications/phenoxyethanol_en ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

4. Cosmetic Ingredient Review. Final report on the safety assessment of Benzyl Alcohol, Benzoic Acid, and Sodium Benzoate. Int J Toxicol. 2001. PMID 11766131. doi:10.1080/10915810152630729 ↩ ↩² ↩³

5. Cosmetic Ingredient Review. Final Report on the Safety Assessment of Sorbic Acid and Potassium Sorbate. J Am Coll Toxicol. 1988. doi:10.3109/10915818809078711 ↩ ↩²

6. Regulation (EC) No 1223/2009 of the European Parliament and of the Council on cosmetic products, consolidated text of 18 May 2026, Annexes III and V. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:02009R1223-20260518 ↩ ↩² ↩³ ↩⁴

7. Cherian P, Bergfeld WF, Belsito DV, et al. Sodium Dehydroacetate and Dehydroacetic Acid. Int J Toxicol. 2024. PMID 38913451. doi:10.1177/10915818241260276 ↩ ↩²

8. Puschmann J, Herbig ME, Müller-Goymann CC. Correlation of antimicrobial effects of phenoxyethanol with its free concentration in the water phase of o/w-emulsion gels. Eur J Pharm Biopharm. 2018. PMID 30118753. doi:10.1016/j.ejpb.2018.08.007 ↩ ↩² ↩³

9. Langsrud S, Steinhauer K, Lüthje S, et al. Ethylhexylglycerin Impairs Membrane Integrity and Enhances the Lethal Effect of Phenoxyethanol. PLoS One. 2016. PMID 27783695. doi:10.1371/journal.pone.0165228 ↩ ↩² ↩³