The Retinol Sandwich Method: One Layer of Moisturizer Kept Activity, Two Layers Cut It About Threefold

If you have ever read that you should buffer retinol by putting moisturizer on before and after it, you have met the sandwich method. It is repeated everywhere, and it sounds sensible: more cushioning, fewer dry, tight evenings. But a 2025 laboratory poster suggests the double layer comes with a trade-off, and the research on what actually keeps retinol comfortable points somewhere less obvious. This guide separates what was measured from what is folklore.

What the sandwich method claims

The classic sandwich is three steps: moisturizer, retinol, moisturizer again. The idea is that the first layer slows how much retinol reaches skin and the second seals everything in. Retinol is a vitamin A derivative that skin converts in steps toward the form that signals to cells, so the amount that gets in and gets converted matters for both results and irritation. The question is whether the buffering costs you anything.

The 2025 poster: one layer kept the signal, two layers dimmed it

At the 2025 American Academy of Dermatology meeting, a team presented an ex vivo study on skin explants from abdominoplasty tissue. They applied a 0.1% retinol cream and a 0.025% tretinoin cream for 48 hours, alone or paired with a lightweight water gel or water cream moisturizer, and then measured gene expression linked to retinoid activity (HBEGF).

The findings were specific. A full sandwich (moisturizer, retinoid, moisturizer) reduced retinoid bioactivity, p<0.05, roughly three-fold, which the presenters attributed to dilution and reduced penetration. An open sandwich, meaning moisturizer on one side only, either before or after, kept bioactivity comparable to the retinoid alone, in either order.

Three caveats matter before you rewrite your routine:

• The tissue came from one donor, with four biopsies per treatment, and it was an ex vivo model, not people using a product for weeks.
• The presenter is head of Medical Affairs at Kenvue, which owns the moisturizer brand tested.
• No quantitative tolerability data were reported. The claim that a full sandwich helps new users acclimate is an interpretation, not a measured result.

So the honest reading is narrow: in this model, one moisturizer layer did not blunt the retinoid, and two layers did.

What is known about moisturizer and comfort

Does adding a moisturizer actually make a retinoid easier to live with? A 2012 randomized, investigator-blinded split-face study ran tretinoin 0.05% nightly for four weeks and applied a ceramide-precursor SPF 30 moisturizer each morning to one randomly chosen half of the face. Roughly 83% to 86% of participants had some irritation on both sides, but it was predominantly mild on the moisturized side, and the authors concluded the moisturizer improved tolerance. Note the details: the study authors include employees of the moisturizer's maker, the sample size was not stated in the version I read, and this was a prescription retinoid, not an over-the-counter retinol.

Still, it supports a modest point. A well-made moisturizer used consistently is associated with milder irritation. It does not show that the sandwich pattern in particular is needed.

Concentration may matter more than layering

The strongest comfort lever in the published retinol data is not layering but dose. A 2022 study in the International Journal of Cosmetic Science compared 0.3% and 1% retinol in 218 women aged 35 to 70 over six weeks, with use building up to nightly. In the 0.3% group, 69.6% reported no reactions, versus 39.8% in the 1% group. Moderate reactions were reported by 7.8% at 0.3% versus 35% at 1%, and 4 people withdrew from the 0.3% group compared with 23 from the 1% group. The authors also reported that the two strengths were similarly effective at remodeling photodamaged skin over long use, although that conclusion came from a small patch-test arm of five volunteers. The work was funded by a beauty company, whose employees were among the authors.

For a beginner, that is a practical case for starting at a lower strength, rather than a high strength plus extra buffering.

Which companion ingredients have patch-test support

If you want something in your formula or routine to take the edge off, a 2025 letter in the Journal of Cosmetic Dermatology tested 15 soothing or barrier-supporting ingredients paired with a 0.1% retinol serum in a five-day patch test with 21 healthy adults. Six combinations showed no Grade 2 reactions: 2% PLG, 5% ceramides, 3% acetyl glucosamine, 2% panthenol, 0.2% TECA and 0.5% Centella asiatica. The retinol serum alone had a mean cumulative irritancy index of 0.77, against 0.26 for the best performer.

Ceramides and panthenol are common, inexpensive moisturizer ingredients, which is the useful part. The limits are real, though: 21 young adults, one population, and a study funded by a cosmetics company whose authors all work there. The authors themselves call for larger trials.

A routine that fits what the evidence supports

None of this is medical guidance, and anyone using a prescription retinoid should follow their prescriber's directions. For over-the-counter retinol, the research above supports a straightforward plan:

1. Pick a low strength first. 0.3% was better tolerated than 1% in the largest comparison here.
2. Ramp up. Start with a couple of nights a week and build to nightly over several weeks, as the tolerance study did.
3. Use one moisturizer layer, not two. Either apply it a few minutes before the retinol or after it. In the explant model, a single layer in either order kept activity comparable.
4. Look for barrier ingredients in that moisturizer. Ceramides and panthenol performed well in the patch test and are easy to find.
5. Use a sunscreen in the morning. The split-face trial used a moisturizer that doubled as an SPF 30, which keeps the routine simple.
6. Judge by how skin looks and feels, not by a rule. Persistent redness, stinging or flaking means step back a notch in strength or frequency.

The takeaway

The sandwich method is popular, but the only published data on it so far suggest the second moisturizer layer may dim what the retinol is there to do, and that evidence is a small, industry-linked laboratory study. The better-supported levers are a lower starting concentration, a gradual ramp, a single supportive moisturizer layer and sunscreen by day. Treat the sandwich as optional, not a requirement.