StratumSkin science, cited
BIngredients·9 min·7 sources

Tretinoin vs Retinol vs Adapalene vs Bakuchiol

Four products claim the same job description; only one of them has forty years of receipts, and it's not the one your influencer is using.

Disclosure

Nothing here is sponsored and we earn nothing from anything mentioned. If that ever changes, this box will say so before you read a word of the article.

Evidence at a glance

Acne evidence strength
A Adapalene and tretinoin have been tested against each other in hundreds of patients across multiple trials with near-identical results.
Anti-ageing evidence strength
B Tretinoin has the deepest trial record, retinol has one solid independent RCT, and bakuchiol has exactly one industry-adjacent comparison.
Irritation risk
B Tretinoin edges out adapalene, retinol and bakuchiol for redness and peeling in head-to-head data, but that ranking rests on a handful of trials, not dozens.
Pregnancy safety
B Pooled first-trimester exposure data looks reassuring, but is too underpowered to overturn standard advice to stop before conceiving.
"Natural alternative" claim (bakuchiol)
C The foundational mechanism study was co-authored by the founder of the company that patented and sells the ingredient.
Time to visible results
C Eight to twelve weeks minimum for any of the four — nobody's skin turns over on a marketing timeline.

Four products, one job description: make your skin behave like it's five years younger, and stop having opinions in the form of acne. Tretinoin, adapalene, retinol and bakuchiol all get filed under "retinoid" in casual conversation, which is a little like filing a Fiat and a Ferrari under "car with wheels." They aim at the same target — the retinoic acid receptors in your skin cells — but they get there by very different routes, backed by very different amounts of proof. Here's what actually separates them, and which one deserves your money first.

What you're actually putting on your face

Tretinoin is retinoic acid itself — the fully active molecule, no conversion required. It's prescription-only everywhere that regulates it seriously, including India, and it's the form most acne and photoaging research has been built on for over 40 years. Adapalene is a synthetic retinoid designed to bind selectively to one receptor subtype instead of all three; it went over-the-counter in the US in 2016, but in India it remains a prescription drug, sold under brands like Deriva and Adaferin. Retinol is vitamin A alcohol — it has to be converted by skin enzymes into retinaldehyde, then into retinoic acid, before it does anything. Two extra steps, each one leaking potency, which is why a 1% retinol serum is nowhere near as strong as a 0.025% tretinoin cream despite the bigger number on the label. Bakuchiol isn't a retinoid at all. It's a compound extracted from babchi seeds (Psoralea corylifolia) that happens to switch on some of the same downstream genes as retinoic acid, without touching the same receptors to get there. "Plant-based retinol alternative" is a marketing phrase built on real but limited biology.

How strong is the evidence, actually

Tretinoin: the one with the paper trail

Tretinoin has the longest and deepest trial record of the four by a wide margin. A 2025 systematic review pooled eight randomized controlled trials in 1,361 patients and found consistent, statistically significant improvement in both fine and coarse wrinkles versus vehicle . That's about as close to settled as topical dermatology gets.

Adapalene: tretinoin's calmer sibling

Adapalene's acne evidence is almost as strong, largely because most of it comes from being tested directly against tretinoin. A meta-analysis pooling five randomized trials in 900 patients found adapalene 0.1% gel matched tretinoin 0.025% gel on lesion reduction, with faster onset in week one and consistently better local tolerability . A separate 168-patient trial comparing adapalene against tretinoin microsphere gel found the same shape of result: comparable efficacy, but significantly better scores for erythema, scaling, dryness, itching and burning with adapalene . Worth flagging: most of this evidence comes from trials run or co-authored by Galderma, the company that makes adapalene. The conclusions have held up in independent reviews since, but it isn't blind, third-party data. What adapalene doesn't have is much independent anti-ageing trial evidence of its own — its wrinkle-fighting reputation mostly rides on being "basically tretinoin."

Retinol: one good trial doing a lot of heavy lifting

Nearly every "retinol is clinically proven" claim you'll see traces back to a small number of actual trials, and one gets cited constantly: a University of Michigan-led, NIH-funded study that applied 0.4% retinol to sun-protected but naturally aged skin in elderly nursing-home residents — 36 enrolled, 23 completed — over several months. Retinol significantly outperformed vehicle on fine wrinkling, roughness and overall severity scores, and increased markers of new collagen production on biopsy . It's a well-designed, independently funded trial — genuinely reassuring. It is also, largely, one trial, on one concentration, in one population. That's a real result, not proof that every retinol serum on a shelf does the same thing at whatever percentage the label declines to disclose.

Bakuchiol: the "natural alternative" with a one-paper reputation

The claim that bakuchiol works "as well as retinol" rests almost entirely on a single randomized, double-blind trial: 44 people split into a bakuchiol-cream group and a retinol-cream group, followed for 12 weeks. Both groups saw significant, statistically similar improvements in wrinkle surface area and pigmentation — and the retinol group reported significantly more scaling and stinging at every check-in . That's a genuinely good result for bakuchiol. It just hasn't been independently repeated. The biology behind why bakuchiol should work at all comes from an earlier paper showing it switches on retinol-like gene expression in cultured skin models — a real finding, but a cell-culture one, never tested on an actual person in that paper. It's also worth knowing who wrote it: the lead author founded the company that patented and sells bakuchiol as a cosmetic ingredient. That doesn't make the chemistry wrong. It does mean the entire "proven natural alternative" pitch is built on one company-linked lab study plus one 44-person trial — not the mountain of independent evidence retinol and tretinoin have behind them.

Bakuchiol isn't a scam. It's just resting a very confident marketing claim on roughly the same amount of clinical evidence you'd want before trusting a bridge.

Irritation, ranked

Line them up by how likely they are to make your face flake. Tretinoin causes the most redness, dryness and peeling because it binds every retinoic acid receptor at full strength with no conversion delay. Adapalene, selective for one receptor, causes measurably less irritation at matched efficacy in head-to-head trials . Retinol is gentler again, partly because a meaningful fraction of what you apply never finishes converting into anything active. Bakuchiol, in its one trial, caused the least reported irritation of the four. None of this means any of them are irritation-free — "less irritating than tretinoin" is a low bar, not a promise of comfort.

Pregnancy and the fine print

All four get the same blanket advice: stop before you're trying to conceive, and don't use any of them while pregnant. For topical retinoids generally, a pooled analysis of pregnancies with first-trimester exposure did not find a clear increase in major birth defects above the background rate — genuinely reassuring on the numbers. But the studies behind that pooled analysis were small, inconsistent in dosing and exposure timing, and nowhere near large enough to rule out a smaller real risk. That's why dermatology guidance hasn't budged: reassuring data isn't proof of safety, and this isn't a decision worth gambling on. Bakuchiol has no dedicated pregnancy-safety studies at all — "natural" is not the same claim as "studied."

Which one should you actually start with

If you've never used a retinoid: start with an over-the-counter retinol, two to three nights a week, buffered under moisturizer. It's the gentlest on-ramp with real, if limited, trial support behind it.

If you have acne and want a prescription option: adapalene is the more sensible first prescription retinoid for most people — matched efficacy to tretinoin with a consistently better tolerability record.

If you've built up tolerance, or your acne or photoaging hasn't responded to anything gentler: tretinoin has the deepest evidence base of the four and is worth the retinisation period it demands.

If you're pregnant, breastfeeding, trying to conceive, or have genuinely tried and can't tolerate any of the above: bakuchiol is a reasonable substitute to discuss with a dermatologist — just don't expect it to be interchangeable with retinol on the evidence, only on the marketing.

The short version

Tretinoin has the most evidence and the most irritation. Adapalene gets you most of tretinoin's results with less drama, but leans on industry-run trials. Retinol's reputation rests on real but sparse independent research. Bakuchiol's "clinically proven" pitch is built on one trial and one company-linked lab study — promising, not proven. Start gentle, escalate only if you actually need to, and don't let a wellness aisle convince you that "natural" and "evidenced" mean the same thing.

This is general education, not medical advice. Nothing here accounts for your history, your medication or your skin, and none of it replaces assessment by a qualified clinician. Where we describe what research found, that is a description of a study — not a promise about you.

References

  1. Huang HY, Lee LTJ Tretinoin for Photodamaged Facial Skin: Systematic Review and Meta-Analysis of Randomized Controlled Trials Dermatology Practical & Concept, 2025.ASystematic review and meta-analysis of 8 RCTs, n=1,361·Authors affiliated with Taipei Medical University Hospital; no industry sponsor identified. Reported significant reductions in fine wrinkles and coarse wrinkles vs vehicle.PubMed 41236273
  2. Cunliffe WJ, Poncet M, Loesche C, Verschoore M A comparison of the efficacy and tolerability of adapalene 0.1% gel versus tretinoin 0.025% gel in patients with acne vulgaris: a meta-analysis of five randomized trials British Journal of Dermatology, 1998.AMeta-analysis of 5 pooled RCTs, n=900 (450/arm), investigator-blinded, intention-to-treat·Poncet, Loesche and Verschoore are Galderma (adapalene's manufacturer) affiliates — industry-run analysis of the company's own trials. Findings of equivalent efficacy with better adapalene tolerability have since been echoed in independent reviews.PubMed 9990421
  3. Thiboutot D, Gold MH, Jarratt MT, Kang S, Kaplan DL, Millikan L, Wolfe J, Loesche C, Baker M Randomized controlled trial of the tolerability, safety, and efficacy of adapalene gel 0.1% and tretinoin microsphere gel 0.1% for the treatment of acne vulgaris Cutis, 2001.BMulticenter RCT, n=168, 12 weeks·Two of nine authors (Loesche, Baker) are Galderma-affiliated. Adapalene and tretinoin were comparably effective; adapalene had significantly fewer treatment-related adverse events and better scores for erythema, scaling, dryness, itching and burning.PubMed 11845943
  4. Kafi M, Kwak HS, Schumacher WE, Cho S, Hanft VN, Hamilton TA, King AL, Neal JD, Varani J, Fisher GJ, Voorhees JJ, Kang S Improvement of naturally aged skin with vitamin A (retinol) Archives of Dermatology, 2007.BRandomized, double-blind, vehicle-controlled paired-skin-site trial; n=36 enrolled (23 completed), mean age 87·Independently funded (NIH K24, Babcock Endowment, Merck-AFAR fellowship) — no cosmetic-industry sponsor. Small elderly sample; single trial at this exact concentration, not yet independently replicated.PubMed 17515510
  5. Dhaliwal S, Rybak I, Ellis SR, Notay M, Trivedi M, Burney W, Vaughn AR, Nguyen M, Reiter P, Bosanac S, Yan H, Foolad N, Sivamani RK Prospective, randomized, double-blind assessment of topical bakuchiol and retinol for facial photoageing British Journal of Dermatology, 2019.BRandomized, double-blind, parallel-group RCT, n=44 (22/arm), 12 weeks·UC Davis-led academic trial. It remains the only published human RCT comparing the two directly — a good result, but a singular one.PubMed 29947134DOI 10.1111/bjd.16918
  6. Chaudhuri RK, Bojanowski K Bakuchiol: a retinol-like functional compound revealed by gene expression profiling and clinically proven to have anti-aging effects International Journal of Cosmetic Science, 2014.DIn vitro gene-expression/DNA microarray study on cultured full-thickness skin models; no controlled human trial despite the title's "clinically proven" claim·Lead author is a founder of Sytheon Ltd, the company that commercializes bakuchiol as a cosmetic ingredient (Sytenol A). This is the foundational paper nearly every 'bakuchiol behaves like retinol' marketing claim traces back to, and its actual data never left the petri dish.PubMed 24471735DOI 10.1111/ics.12117
  7. Kaplan YC, et al. Pregnancy outcomes following first-trimester exposure to topical retinoids: a systematic review and meta-analysis British Journal of Dermatology, 2015.ASystematic review and meta-analysis of pregnancy cohort/registry data on first-trimester topical retinoid exposure·Academic/MotherToBaby-affiliated authorship, not industry-funded. Found no statistically significant increase in major malformations above baseline, but underlying studies were small and heterogeneous — standard clinical advice to avoid use in pregnancy has not changed as a result.PubMed 26215715

Your turn

Did this hold up? Tell us honestly — especially if it didn’t.

Anonymous. Stored in your browser.

The Sunday study

One paper a week, explained properly

We pick a single new study, tell you what it measured, how many people it measured it in, and whether it should change anything you're doing — plus the two articles we published that week. No panic, and no hype.

We don't recommend products yet. When we do, they'll be things you can actually buy in India, any affiliate link will be labelled as one, and a commission will never buy a recommendation.

One email each Sunday. Unsubscribe in one click.