How Much Retinol Is Actually in the Product, and How Much Survives Six Months?

Less than the label states, in a measurable number of cases — and in storage trials a large fraction is gone within months. Across 12 commercial cosmetics followed for six months, retinoid content declined by 0–80% at 25 °C and by 40–100% at 40 °C. In a separate survey of commercial products, measured content deviated from the declared figure in seven products, across a 0–130% range.

Published 3 November 2026. Last reviewed 3 November 2026.

Two questions that are usually merged

A declared percentage makes two separate promises: that the stated amount was present when the product was filled, and that it is still present when the product is used. Both have been tested analytically, and both fail in a measurable proportion of commercial products.

The terms used below, each defined to stand alone. A retinoid is any member of the vitamin A family of compounds, natural or synthetic. Retinol is the alcohol form, the one most often declared in cosmetics. Retinyl palmitate is its fatty-acid ester, more stable and requiring conversion. Hydroxypinacolone retinoate is a newer ester of retinoic acid. Retinoic acid is the acid form, which is a medicinal substance and is prohibited in cosmetic products. First-order kinetics means a decay in which the rate of loss at any moment is proportional to the amount still present, so that a constant fraction — not a constant quantity — is lost per unit time. Photostability is resistance to chemical change on exposure to light.

What was measured at the point of sale

Temova Rakuša, Škufca, Kristl and Roškar analysed commercial retinoid cosmetics by validated HPLC-UV, with LC-MS confirmation of hydroxypinacolone retinoate and a dedicated screen for retinoic acid. Retinol and retinyl palmitate were the forms most often found, “in concentrations mostly below 0.3%, and up to 1.3% retinol equivalents”. Measured contents “deviated significantly from the quantitatively declared ones in seven products (0%–130%)”, and “in more than half of the tested products, inconsistencies between the contained and labeled retinoid were noticed” — that is, the form declared was not the form present. Retinoic acid, which is not permitted in cosmetics, was detected in two products, and 20% of the tested cosmetics exceeded the recommendations of the Scientific Committee on Consumer Safety (J. Cosmet. Dermatol. 2021;20(4):1166–1175).

What was left after six months

The same group then followed stability. Sixteen retinoid derivatives across 12 commercial cosmetics were monitored through six months of long-term and accelerated testing, plus a one-week photostability study. Degradation followed first-order kinetics, which allowed shelf-life prediction. The outcome was formulation-dependent and, in most products, substantial.

Declared content, measured content and measured loss in commercial retinoid cosmetics

Condition	Period	Reported change in retinoid content	Material tested	Source
Long-term storage, 25 °C	6 months	0%–80% decline	16 retinoid derivatives in 12 commercial cosmetics	Temova Rakuša et al. 2021b
Accelerated storage, 40 °C	6 months	40%–100% decline	Same set	Temova Rakuša et al. 2021b
Photostability study	1 week	Light degradation “more pronounced than temperature-induced degradation”; no single percentage reported	Same set	Temova Rakuša et al. 2021b
Measured vs. quantitatively declared content	As purchased	Significant deviation in 7 products, spanning 0%–130%	Commercial retinoid cosmetics	Temova Rakuša et al. 2021a
Tretinoin in methanol, 30 W lamp at 366 nm*	1 h	27% of the initial amount remaining; half-life about 0.4 h	Solution	Kryczyk-Poprawa et al. 2020
Tretinoin in nanosuspension, same lamp*	1 h	83% remaining; half-life about 3 h	Nanosuspension	Kryczyk-Poprawa et al. 2020

* The last two rows concern tretinoin, a prescription retinoid, not retinol. They are included because they quantify the effect of the carrier on photostability; the magnitudes are not transferable to retinol.

Light does more than heat

The storage study's own conclusion on this point is explicit: “Light degradation was more pronounced than temperature-induced degradation.” The photochemistry behind that is partly counter-intuitive. Kryczyk-Poprawa, Kwiecień and Opoka record that vitamin A “absorbs ultraviolet radiation with an absorption maximum of λ = 325 nm”, and that for tretinoin “the most harmful wavelength was about 420 nm, which directly contributes to the photodegradation of tretinoin, not 350 nm—the wavelength of maximum absorption”. They also identify UVA (315–400 nm) as “the major contributor to the photodegradation of tretinoin and isotretinoin” (Pharmaceutics 2020;12(1):10).

Two limits on that evidence deserve stating rather than glossing. First, the wavelength of maximum absorption is not necessarily the wavelength of maximum damage, so an absorption spectrum alone does not predict shelf behaviour. Second, and more importantly, nearly all the quantified photodegradation figures in that review concern tretinoin and isotretinoin. For retinol and retinyl palmitate the review reports no half-life, no percentage and no dose — only the absorption maximum, and a qualitative observation that photostability was higher in formulations containing UV filters. The quantitative picture for cosmetic retinol rests on the storage studies above, not on photochemical half-lives.

What has been shown to slow the loss

The carrier matters more than any single additive. In the review's compiled figures, moving tretinoin from a methanolic solution into a nanosuspension raised the fraction surviving one hour of 366 nm irradiation from 27% to 83%, and the half-life from about 0.4 h to about 3 h; a nanoemulsion gave 52% and about 0.9 h. Micronised tretinoin in a 0.05% gel degraded by 11–12% over 8 h of irradiation, against 85–90% under fluorescent light and 84–89% under simulated sunlight for a conventional gel. An isotretinoin micro-emulsion retained 75% of the initial concentration after 240 minutes while the methanolic solution degraded completely.

Two reported failures are as informative. Adding a surfactant or using β-cyclodextrin “did not improve its photostability” for tretinoin in a lotion, and the solubilising agent Brij 35 S had a negative influence on tretinoin stability in a Carbopol gel. Encapsulation is not automatically protective either: Chmykh and Nadeau report that the gain from liposome encapsulation of retinol “is limited because of lipid peroxidation”, with antioxidants — butylated hydroxytoluene and a proprietary mixture — improving shelf stability, and the proprietary mixture protecting retinol “for significantly longer than BHT and via different mechanisms” under UV irradiation (ACS Omega 2020;5(29):18367–18375). The vesicle that carries the active can itself become the oxidising environment.

Where the regulatory frame sits

In the European Union, vitamin A in cosmetic products is restricted under Regulation (EC) No 1223/2009, as amended by Commission Regulation (EU) 2024/996 of 3 April 2024 — “amending Regulation (EC) No 1223/2009 … as regards the use of Vitamin A, Alpha-Arbutin and Arbutin and certain substances with potential endocrine disrupting properties in cosmetic products” — which entered into force on 1 May 2025. The permitted concentrations and the product categories they apply to are set out in the Annex III entry for vitamin A and should be read in the current consolidated text (EUR-Lex, CELEX:32024R0996). Retinoic acid remains outside the cosmetic category altogether.

The analytical findings and the legal limits interact in a way that is easy to miss. A compliance figure is a figure at a moment in time. If a product that complies at filling loses 80% of its retinoid in six months at room temperature, the declared percentage has stopped describing the product long before the container is empty — and if it is overloaded to compensate, it may not have complied at filling. Both failure modes appear in the data above.

What follows for formulation and specification

The authors of the stability study draw the operational conclusion themselves: retinoid “chemical stability and physical stability in topical formulations need to be evaluated by real-time stability studies, instead of the more frequently used accelerated stability studies”. The 40 °C and 25 °C figures above are not interchangeable, and a six-month accelerated result does not substitute for a six-month real one.

A specification that can be audited therefore states four things rather than one: the retinoid form actually present, confirmed analytically; its content at release; its measured content at the end of the declared period under real-time conditions; and the protective strategy — carrier, antioxidant system, packaging — with evidence that it holds. Formulating to the criteria of compounding pharmacy means committing to the fourth column, not only the first. The same reasoning applies to ascorbate esters, where oxidative stability rather than declared concentration governs what remains available; TTS Clinical sets out that case in an open-access technical report (doi:10.5281/zenodo.22491021).

References

1. Temova Rakuša Ž., Škufca P., Kristl A., Roškar R. (2021a). Quality control of retinoids in commercial cosmetic products. Journal of Cosmetic Dermatology 20(4):1166–1175. doi:10.1111/jocd.13686 · PMID 32813932
2. Temova Rakuša Ž., Škufca P., Kristl A., Roškar R. (2021b). Retinoid stability and degradation kinetics in commercial cosmetic products. Journal of Cosmetic Dermatology 20(7):2350–2358. doi:10.1111/jocd.13852 · PMID 33206444
3. Kryczyk-Poprawa A., Kwiecień A., Opoka W. (2020). Photostability of topical agents applied to the skin: a review. Pharmaceutics 12(1):10. doi:10.3390/pharmaceutics12010010
4. Chmykh Y. G., Nadeau J. L. (2020). Characterization of retinol stabilized in phosphatidylcholine vesicles with and without antioxidants. ACS Omega 5(29):18367–18375. doi:10.1021/acsomega.0c02102
5. Commission Regulation (EU) 2024/996 of 3 April 2024 amending Regulation (EC) No 1223/2009 as regards the use of Vitamin A, Alpha-Arbutin and Arbutin and certain substances with potential endocrine disrupting properties in cosmetic products. EUR-Lex, CELEX:32024R0996
6. TTS Clinical Scientific Team (2026). Topical vitamin C: the pH constraint on L-ascorbic acid and the oxidative stability of tetrahexyldecyl ascorbate, v1.0. doi:10.5281/zenodo.22491021

Scope note. This article describes analytical and stability findings published in the cited scientific literature, and the regulatory instruments named. It is not a product claim, and nothing in it should be read as a statement about the effects of any TTS Clinical or Truth Treatment Systems product, or as medical advice. Figures obtained with prescription retinoids are reported as such and do not transfer to cosmetic retinoids. Concentration limits and product categories must be read in the current consolidated text of the applicable regulation.