Tretinoin Sandwich Method: What Has Actually Been Tested?
Start with our pillar guide: What Is the Skin Barrier? The Science of Hydration, Ceramides and Occlusives, or browse all Skincare Fundamentals articles.
Evidence label: Very limited evidence for the exact sequence. Limited, partly indirect evidence for adjunct moisturizer. The tretinoin sandwich method usually means moisturizer applied on both sides of a topical tretinoin layer. In the PubMed records we screened, we found no controlled trial comparing that before-and-after sequence with standard tretinoin use. A few tretinoin studies support using moisturizer alongside treatment for comfort. None of the studies reviewed shows that a full sandwich preserves absorption, acne results or wrinkle benefit.
This page describes research. It is not a routine. If you have questions about how moisturizer fits with a prescribed tretinoin product, ask the clinician who prescribed it.
What does the tretinoin sandwich method mean?
Evidence label: terminology only, not efficacy evidence.
People use several labels online, and they describe different things:
- Closed sandwich: moisturizer, then tretinoin, then moisturizer again.
- Open sandwich or buffering: moisturizer applied before tretinoin only.
- Preconditioning: moisturizer used for a period before tretinoin is started.
- Adjunct moisturizer: moisturizer used during treatment, with no specified order.
These are separate exposures. A study of one does not automatically answer questions about another. Retinol frequency and general tolerance are covered in Can I use retinol every night?. General routine order is covered in our skincare protocol. This page stays tretinoin-specific.
Has the tretinoin sandwich method been tested directly?
Evidence label: Very limited evidence. No direct closed-sandwich comparison was identified in these searches.
On 8 November 2026 we searched PubMed titles and abstracts for sandwich, buffering or layering terms combined with tretinoin or retinoid. The search returned 16 records. On screening, none was a relevant comparison of sandwiching facial tretinoin. A second search, for tretinoin plus moisturizer or preconditioning, returned 13 records. Those included the only order-specific report we found.[^1] We also ran a broader search on adjunct tolerability for context.
We did not search Embase, CENTRAL, trial registries or unpublished protocols, so we cannot rule out a trial outside these searches. Material described online as "open sandwich research" is not automatically an indexed randomized trial. We excluded off-topic results about molecular buffering and retinal proteins. None of the 11 selected records carried a retraction or expression-of-concern flag.
Evidence map
| Claim or use | What was actually tested | What cannot be inferred | Evidence label |
|---|---|---|---|
| Moisturizer before AND after tretinoin (closed sandwich) | No relevant controlled sequence comparison identified in the specified searches | Reduced irritation, unchanged absorption or equal long-term efficacy for that sequence | Very limited evidence |
| Moisturizer BEFORE tretinoin (open sandwich/buffering) | One sparse acne report of moisturizer-before use in a multidrug regimen[^1] | That the order caused any benefit, or that penetration is unchanged | Very limited evidence |
| Preconditioning before starting tretinoin | Randomized two-moisturizer side comparison, with tretinoin introduced later[^2] | Immediate layering benefit, or any recommended lead-in schedule | Limited |
| Adjunct moisturizer during tretinoin | Split-face and half-face comfort studies[^3][^4] | That any moisturizer, any skin type or injured skin behaves the same | Limited |
| Moisturizer with adapalene or adapalene/BPO | Five trials with different drugs and vehicles[^5][^6][^7][^8][^9] | Tretinoin-sandwich effects | Limited, indirect |
| Branded tretinoin lotion vs generic cream | Two-week vehicle comparison[^10] | Effects of a separate moisturizer layer or its order | Early-preliminary |
What do tretinoin moisturizer studies show?
Evidence label: Limited, partly indirect to the sandwich sequence.
Moisturizer before tretinoin, in a combination regimen.[^1] This acne report applied a ceramide-containing moisturizer before each topical treatment, including evening tretinoin, and reported safety and efficacy. The abstract gives no sample size, design, comparator or numerical outcomes. The order is explicit. But the report cannot separate the effect of the order from the rest of the regimen. Disclosure: Mount Sinai School of Medicine affiliation. PubMed indexes no grant or COI statement, which is not proof that there was none. We did not check the full text.
Preconditioning.[^2] In this paired, double-blind, randomized study, women with facial photodamage applied a different moisturizer to each side of the face for 10 weeks. One moisturizer contained niacinamide, panthenol and tocopheryl acetate. The other had similar moisturizing ingredients but no vitamins. Full-face tretinoin began after an initial moisturizer-only period. The barrier-enhancing side showed easier early retinization and an augmented treatment response. The abstract does not give a sample size. The study compared two moisturizers in a preconditioning design. It did not compare moisturizer with no moisturizer, and it did not test before-and-after layering. Disclosure: Wake Forest University School of Medicine affiliation. No grant or COI statement is indexed. Full text not checked.
Adjunct moisturizer, healthy skin.[^3] This was a randomized, investigator-blinded, split-face study. Participants applied tretinoin to the whole face and a moisturizer with SPF to one side. About 83–86% had irritation on both sides, but it was mostly mild on the moisturizer side. Tolerability preferences favored that side. The abstract does not state the sample size, duration or application order. Disclosure: TKL Research, Inc. affiliation. The product tested was a proprietary Cetaphil moisturizer. Manufacturer funding is not verified, no COI statement is indexed, and we did not check the full text.
Adjunct moisturizer, mixed retinoid users.[^4] Thirty people using either oral isotretinoin or topical tretinoin applied a moisturizing cream to one half of the face for 15 days and left the other half untreated. Dryness, roughness, scaling and discomfort improved. The abstract does not give the size of the tretinoin subgroup and reports no acne efficacy endpoint. This page offers no guidance on oral retinoids. Disclosure: PERITESCO (Paris) affiliation. The product was a branded Cetaphil cream. No grant or COI statement is indexed. Full text not checked.
Overall, moisturizer used alongside tretinoin was generally linked with better comfort. None of these studies tested whether applying moisturizer both before and after tretinoin changes the outcome.
Do adapalene studies prove the sandwich method works for tretinoin?
Evidence label: Limited, indirect.
Adapalene is a different retinoid, and several of these trials combined it with benzoyl peroxide (BPO). They show that moisturizers helped tolerability with those products. They do not show the same for a tretinoin sandwich.
- Healthy volunteers.[^5] A randomized, evaluator-blinded, split-face trial in 30 healthy Chinese volunteers lasted four weeks. The global worst score was lower on the moisturizer side (0.43 vs 0.59, P=.032). Disclosure: Fujita Health University affiliation. No grant or COI statement is indexed. Full text not checked.
- Acne, vehicle-controlled.[^6] In this double-blind trial, 120 acne patients were split into three groups for eight weeks. The active-moisturizer group improved in lesions compared with baseline and did better than the other groups on instrument-measured tolerability. Disclosure: Siriraj Hospital, Mahidol University affiliation. No grant or COI statement is indexed. Full text not checked.
- Adapalene/BPO regimens.[^7] This multicenter, single-blind trial randomized 120 people to four early regimens. The moisturizer regimen prevented dryness and scaling. Lesion reductions at week 12 were similar across groups, but similar results are not a formal noninferiority finding. Disclosure: Authors include Nestlé Skin Health–Galderma R&D and Galderma International affiliations, and Galderma makes the gel studied. No grant or COI statement is indexed. Full text not checked.
- Ceramide/niacinamide vs hydrophilic cream.[^8] This eight-week, double-blind, split-face trial enrolled 40 people. Lesion outcomes favored the ceramide formula. Barrier and tolerance measures improved, but did not differ significantly between the two treatments. Disclosure: University, hospital and clinic affiliations in Thailand and Indonesia. Accessible full text reports no funding for the article and no conflicts of interest; methods name a CeraVe PM moisturizer (Valeant).
- Cleanser plus lotion.[^9] This double-blind trial compared a ceramide cleanser and lotion with a basic face wash. At week 12 the ceramide group had lower water loss, less dryness and fewer inflammatory lesions. Because two products were used together, the benefit cannot be credited to the moisturizer alone. Disclosure: Affiliations include Dermatology Consulting Services and CeraVe (New York). An unrestricted educational grant from CeraVe, LLC (a L'Oréal division) supported the research; two authors are CeraVe employees. Full text reports 100 enrolled and 91 completers (45 ceramide regimen, 46 control).
Does the sandwich method make tretinoin less effective?
Evidence label: Very limited evidence. The question is unresolved for the exact sequence.
None of the selected records measured tretinoin absorption under before-and-after layering. None compared acne or wrinkle outcomes between a sandwich and standard use. Less irritation, or similar lesion counts with a different retinoid, cannot settle that question. The preconditioning study reported an augmented response, but it compared two moisturizers.[^2]
So neither claim, that buffering "blocks" tretinoin or that it "doesn't dilute" it, is supported here. Long-term wrinkle evidence is covered in Does tretinoin help with wrinkles?. Acne tolerability findings should not be read as preserved wrinkle efficacy.
Does the tretinoin formulation itself matter?
Evidence label: Early-preliminary, formulation-specific.
In a single-center, double-blind, split-face trial, 25 adult women with acne applied a branded tretinoin lotion to one cheek and a generic cream to the other for two weeks.[^10] The cream side had significantly more redness, scaling and dryness, and about 70% of participants preferred the lotion. This is a comparison of vehicles, not of an external moisturizer layer. Disclosure: One co-author is affiliated with Ortho Dermatologics (a Bausch Health division), which markets the branded lotion tested (Altreno). A separate study-funding statement was not verified beyond that public author page; PubMed indexes no grant list or COI statement.
An older overview reported that pairing tretinoin with compatible moisturizers, cleansers and sunscreens improved tolerance, but it did not report the number of participants.[^11] It is background only, not a sandwich trial. Disclosure: Neutrogena Corporation affiliation. No grant or COI statement is indexed. Full text not checked.
Labels such as "oil-free," "ceramide" or "occlusive" are not trial-proven to be interchangeable. This page does not rank products.
What can't this evidence tell you about irritated or recently treated skin?
Evidence label: applicability gap.
These studies involved intact skin, acne or photodamage. Being tolerated in a trial does not make something suitable during dermatitis, after laser or microneedling, near the eyes, or with every prescription vehicle. None of the studies is a restart protocol, a treatment for a reaction or a reason to add actives.
For background, see how to fix a damaged skin barrier. For post-procedure questions, which are a separate topic, see retinol after microneedling. Neither page replaces a regimen. Ask the prescribing clinician how this evidence relates to your prescribed product.
Where the evidence stops
Evidence label: Limited.
- In the PubMed searches described above, we found no direct comparison of the closed-sandwich sequence.
- Several abstracts omit sample size, duration or comparator.
- Follow-up was short, from 15 days to 12 weeks.
- Many tested formulas were proprietary, and several studies had company affiliations or tested branded products.
- Populations were mixed: healthy volunteers, acne patients and photodamaged skin. Many of the trials used adapalene rather than tretinoin.
- None of the 11 records has an indexed PMC full text. We checked full texts for none of them, so funding and sample-size sections remain unverified. An empty PubMed funding field does not mean a study was unfunded or free of conflicts.
FAQ
What is the tretinoin sandwich method?
It is an informal term, usually for a layer of moisturizer applied before and after a topical tretinoin layer. Online labels vary. Some people use the term for moisturizer applied before tretinoin only. It is a description of a practice, not an evidence-backed protocol.
Is moisturizer-before the same as a full sandwich?
No. Moisturizer applied only before tretinoin, sometimes called buffering or an open sandwich, is a different order from moisturizer applied both before and after. The one order-specific report we found tested moisturizer-before within a multidrug acne regimen, not a full sandwich.
Has sandwiching been shown to reduce tretinoin irritation?
Not for the exact before-and-after sequence in the PubMed searches reviewed here. Some tretinoin studies found better comfort when a moisturizer was used alongside treatment. They did not test whether sandwiching specifically caused that benefit.
Does sandwiching make tretinoin less effective?
The reviewed studies do not answer this. None measured tretinoin absorption or compared acne or wrinkle outcomes between a full sandwich and standard use. Claims that it blocks tretinoin, or that it does not dilute it, both go beyond this evidence.
Do adapalene moisturizer trials prove it works for tretinoin?
No. Adapalene is a different retinoid, and several of these trials used adapalene combined with benzoyl peroxide in different vehicles. They show moisturizers helped tolerability with those products, which is indirect evidence for tretinoin.
Does every moisturizer or tretinoin vehicle behave the same?
That has not been established. The studies tested specific, often proprietary formulas, and one small split-face trial found two tretinoin vehicles differed in tolerability. Ingredient labels such as ceramide or oil-free are not proven interchangeable.
Can intact-skin studies guide use after microneedling or during a reaction?
No. These studies involved healthy skin, acne or photodamage, not dermatitis or recently treated skin. They are not a tested restart protocol. Questions about irritated or treated skin belong with the prescribing clinician.
Is moisturizer preconditioning the same as sandwiching tretinoin?
No. Preconditioning means using moisturizer for a period before tretinoin is started. One randomized study tested this design and compared two moisturizers. That is not the same as layering moisturizer before and after each tretinoin application, and it is not a recommended lead-in schedule.
References
[^1]: PMID 22648223. Report of a ceramide-containing moisturizer applied before morning and evening topical acne treatments, including tretinoin gel, for facial acne vulgaris. Source scope: PubMed abstract and metadata. Funding/COI: Mount Sinai School of Medicine affiliation; no grant list or COI statement indexed (not proof of none); full text not verified. [^2]: PMID 17121065. Facilitating facial retinization through barrier improvement; paired double-blind randomized moisturizer comparison in photodamaged women. Source scope: PubMed abstract and metadata. Funding/COI: Wake Forest University School of Medicine affiliation; no grant list or COI statement indexed; full text not verified. [^3]: PMID 23135655. Randomized, investigator-blinded split-face study of an adjunctive ceramide-precursor SPF moisturizer with topical tretinoin in healthy skin. Source scope: PubMed abstract and metadata. Funding/COI: TKL Research, Inc. affiliation; proprietary Cetaphil product tested; no grant list or COI statement indexed; manufacturer funding not verified; full text not verified. [^4]: PMID 17373148. Moisturizing cream as adjunctive treatment to oral isotretinoin or topical tretinoin in acne. Source scope: PubMed abstract and metadata. Funding/COI: PERITESCO affiliation; branded Cetaphil cream tested; no grant list or COI statement indexed; full text not verified. [^5]: PMID 22384983. Randomized, investigator-blinded split-face study of a non-comedogenic moisturizer with adapalene gel in healthy Asian subjects. Source scope: PubMed abstract and metadata. Funding/COI: Fujita Health University School of Medicine affiliation; no grant list or COI statement indexed; full text not verified. [^6]: PMID 26293170. Double-blind, randomized, vehicle-controlled study of an anti-inflammatory moisturizer with adapalene gel in acne. Source scope: PubMed abstract and metadata. Funding/COI: Siriraj Hospital, Mahidol University affiliation; no grant list or COI statement indexed; full text not verified. [^7]: PMID 30187864. Randomized controlled study of four adapalene/benzoyl peroxide gel regimens for acne vulgaris. Source scope: PubMed abstract and metadata. Funding/COI: author affiliations include Nestlé Skin Health–Galderma R&D and Galderma International; no grant list or COI statement indexed; full text not verified. [^8]: PMID 38299457. Split-face, double-blind randomized trial of a ceramide/niacinamide moisturizer versus hydrophilic cream with topical anti-acne treatment. Source scope: PubMed abstract, metadata and accessible Wiley OA full text reviewed. Funding/COI: full text reports no funding for the article and no conflicts of interest; methods name CeraVe PM (Valeant); university/clinic affiliations as indexed. [^9]: PMID 37276158. Ceramide-containing adjunctive skin care for skin barrier restoration during acne vulgaris treatment. Source scope: Abstract, PubMed metadata and accessible JDD full text reviewed. Funding/COI: unrestricted educational grant from CeraVe, LLC (L'Oréal division); two authors are CeraVe employees; other author affiliations as in the article. Sample: 100 enrolled, 91 completed. [^10]: PMID 35946981. Vehicle formulation impacts tolerability and patient preference: tretinoin branded lotion versus generic cream. Source scope: Abstract, PubMed metadata and accessible JDD author/affiliation page reviewed. Funding/COI: PubMed grant/COI not indexed; public author list includes Ortho Dermatologics / Bausch Health affiliation for one co-author; full funding paragraph behind login not verified. [^11]: PMID 10393518. Retinoid therapy: compatible skin care. Source scope: PubMed abstract and metadata. Funding/COI: Neutrogena Corporation affiliation; no grant list or COI statement indexed; full text not verified.