Bakuchiol vs Retinol: One Trial, Read Closely

Bakuchiol has the best marketing in skincare and the thinnest evidence. Every bottle calls it a natural retinol alternative with the same results and none of the irritation. The claim rests on one trial, and the trial is worth reading closely, because what it actually found is more interesting than the label.

The one trial that matters

Dhaliwal and colleagues ran the trial in 2019 at UC Davis and published it in the British Journal of Dermatology. Forty-four adults aged 31 to 56 were randomized and double-blind for 12 weeks. One group applied a 0.5 percent bakuchiol cream twice a day. The other applied a 0.5 percent retinol cream every night, the way retinol is normally used. High-resolution facial photography and blinded dermatologist scoring measured wrinkle severity, hyperpigmentation, skin elasticity, and redness at baseline, 4, 8, and 12 weeks.

Both groups improved significantly on wrinkle surface area and hyperpigmentation, and there was no statistically significant difference between them on any efficacy measure. At week 12, wrinkle surface area was down 19.0 percent with bakuchiol and 23.2 percent with retinol. Hyperpigmentation improved for 59 percent of the bakuchiol group and 44 percent of the retinol group, graded by a dermatologist. The numbers favor retinol slightly on lines and bakuchiol slightly on pigment, but neither gap reached statistical significance.

What matched and what did not

Where the two ingredients diverged sharply was tolerability. The retinol group reported significantly more facial scaling and stinging at every check-in, and more redness and skin tightness. The bakuchiol group reported significantly less of both. That is the real finding. Bakuchiol did not beat retinol. It matched retinol on the outcomes that were measured, and it was meaningfully easier to live with.

The limits matter as much as the results. Forty-four people is a small study, it ran for twelve weeks, it had no group using a plain cream, and it has not been independently replicated. The retinol concentration, 0.5 percent, sits at the higher end of over-the-counter retinol but below prescription tretinoin. Twelve weeks may be too short to detect differences in collagen synthesis, which takes longer. Retinol and prescription tretinoin still have a far larger body of evidence for deeper, long-term anti-aging.

Who it is actually for

Bakuchiol is for the person who quit retinol because of irritation, not the person chasing the strongest possible result. If your skin cannot tolerate retinol, bakuchiol at 0.5 percent twice a day is the best-supported alternative, and the trial says you are not giving up much on lines and pigment at twelve weeks. If your skin tolerates retinol, there is no reason in the data to switch. The trial also says something useful about how to use it: twice a day, not once, and give it the full twelve weeks, because at four and eight weeks the changes were small.

Bakuchiol is not a retinoid. It is a meroterpene compound extracted from the seeds of Psoralea corylifolia, the babchi plant, and it is structurally unrelated to vitamin A. It does not bind retinoic acid receptors. That is why it does not irritate like a retinoid, and also why the mechanism behind its effects is still less understood than retinol's.

The pregnancy question it cannot answer

Direct trials of bakuchiol in pregnancy and lactation remain absent. Because it is structurally non-retinoid, the theoretical risk is low, but low theoretical risk is not the same as tested safety. Anyone pregnant, trying to conceive, or breastfeeding should ask a doctor before using it, rather than treating it as a proven safe swap. The honest version of the bakuchiol story is a good one: one small, rigorous trial, comparable results at twelve weeks, clearly better tolerance, and a lot that is still untested.

Frequently asked questions

Is bakuchiol as effective as retinol?

On the two outcomes tested, yes, at 12 weeks. In the Dhaliwal trial, 0.5% bakuchiol matched 0.5% retinol on wrinkle surface area and hyperpigmentation, with no statistically significant difference. That is one small trial, not a body of evidence.

How long does bakuchiol take to work?

Twelve weeks. At four and eight weeks the changes were small. The change that counted came at twelve, with wrinkle surface area down about 19 percent for bakuchiol and 23 percent for retinol.

Is bakuchiol safe during pregnancy?

There is no direct clinical data. Bakuchiol is not a retinoid, so the theoretical risk is low, but no trial has tested it in pregnancy or lactation. Ask your doctor rather than assuming it is safe.

What percentage of bakuchiol should I use?

0.5 percent, twice a day, is the only concentration with clinical data behind it. Serums at 1.25 percent are not proven to outperform 0.5 percent. The trial data is built on the 0.5 percent cream.