Bakuchiol: does the “natural retinol” actually work?

What bakuchiol actually is

Bakuchiol is a meroterpene — a plant compound that mixes a terpene fragment with a phenol — extracted from the seeds of Psoralea corylifolia, a plant known in Ayurvedic and traditional Chinese practice as babchi.⁵ If you look at its chemical structure next to a retinoid, the two share essentially nothing. Retinoids are vitamin-A derivatives; bakuchiol is not vitamin A, not a retinoid, and not a precursor to one. That fact alone should settle the most common piece of marketing shorthand: bakuchiol is not “plant retinol” in any literal sense.

And yet the interest is not invented. The reason bakuchiol broke out of the natural-skincare niche is that, despite the structural mismatch, it behaves in skin in ways that overlap meaningfully with how retinol behaves — which is a far more interesting claim than “it’s a gentle moisturizer.” To understand why a molecule that looks nothing like a retinoid can act a bit like one, you have to look at the signal it pulls inside the cell, not its shape on paper.

The mechanism: retinol’s effect, not retinol’s structure

The foundational work here is a 2014 study by Chaudhuri and Bojanowski that profiled what bakuchiol does to gene expression in skin cells and compared it directly to retinol.³ The headline finding was that, across a DNA-microarray screen, bakuchiol and retinol produced strikingly similar overall gene-expression profiles — the authors described bakuchiol as a “functional analogue” of retinol, a compound that lands on the same downstream switches without sharing the same chemistry.³

Concretely, bakuchiol up-regulated types I and IV collagen in those screens and stimulated type III collagen in mature skin-cell models — the structural proteins whose decline shows up as thinning, sagging, and lines.³ It also raised expression of aquaporin 3, a water-channel protein tied to skin hydration.³ This is the signal it pulls: a retinol-like push on the collagen-building machinery, arrived at through a different molecular door. Where a true retinoid works by binding nuclear retinoic-acid receptors directly, bakuchiol appears to converge on similar gene targets without occupying that receptor — which is exactly why it can produce retinol-like remodeling while side-stepping some of the receptor-driven irritation.

Bakuchiol doesn’t imitate retinol’s molecule. It imitates retinol’s instructions to the cell — reaching a similar collagen signal through a different door, and skipping some of the irritation that comes with the retinoid receptor.

Two honest caveats sit on top of this elegant story. First, gene-expression similarity is mechanism-plausible support, not proof of equal clinical benefit — two compounds can nudge the same genes and still differ in how much visible change they produce on a real face over months. Second, this anchoring mechanistic work was conducted by researchers connected to bakuchiol’s commercial development, so the foundational “retinol-like” framing comes from an interested source. That does not make it wrong — the gene data are the gene data — but it means the cleanest confirmation has to come from the clinical trials, which is where we go next.

The head-to-head evidence

The single most important study in this entire topic — the one cited every time bakuchiol comes up — is Dhaliwal and colleagues’ 2019 trial in the British Journal of Dermatology.¹ It is the rare piece of skincare research that does the hard thing: a prospective, randomized, double-blind comparison of bakuchiol against retinol, head to head, on real faces.

Forty-four participants applied either 0.5% bakuchiol cream twice daily or 0.5% retinol cream once daily for 12 weeks, with a dermatologist blinded to which group was which scoring the results.¹ The outcome: both compounds significantly decreased wrinkle surface area and hyperpigmentation, and — the line that launched a thousand product pages — there was no statistically significant difference between them on those measures.¹ On the things people actually buy a retinoid for, bakuchiol kept pace.

That is a genuinely strong result for a natural-skincare ingredient, and it deserves to be stated plainly rather than hedged into mush: in this trial, bakuchiol did not lose to retinol. But the same plainness demands the limits be stated just as clearly. This was one trial of 44 people over 12 weeks — a small, short study, not a large multi-site or long-duration one. “No significant difference” in a 44-person trial is also partly a statistical-power story: a small study can fail to detect a real gap between two treatments simply because it isn’t big enough to see it. Comparable is the right word; identical is not.

The broader literature is consistent with that single trial without dramatically expanding it. A 2022 systematic review pulled together 30 articles on bakuchiol — the bulk preclinical, with only seven clinical trials among them — and concluded that for photoaging, acne, and post-inflammatory hyperpigmentation, bakuchiol showed “beneficial results comparable to those achieved by topical retinoids,” while explicitly calling for more study to define its real place in dermatology.² That is the honest shape of the evidence: a coherent, positive signal across a handful of small studies, anchored by one good head-to-head, and openly described by its own reviewers as not yet settled.

Why “better tolerated” is the real story

If the efficacy data say bakuchiol can keep pace, the tolerability data are arguably the more decisive reason anyone actually switches. In the Dhaliwal trial, the retinol group reported more facial skin scaling and stinging than the bakuchiol group — the classic “retinization” complaints of dryness, flaking, and burning that drive so many people to abandon retinoids in the first few weeks.¹ The conclusion the authors drew was specific: bakuchiol is comparable to retinol in improving photoaging and is better tolerated.¹

That tolerability theme holds up in the population most likely to struggle with retinoids. A 2020 study put a bakuchiol-based regimen through a four-week trial in 60 women aged 40 to 65 with sensitive skin — one-third each with eczema or atopic dermatitis, rosacea, and cosmetic-intolerance syndrome, precisely the people for whom a standard retinoid is often a non-starter.⁵ The regimen was well tolerated and produced significant improvements in smoothness, clarity, radiance, and overall appearance, with measurable gains in skin moisture.⁵ It is a short, single-arm tolerability study rather than a head-to-head efficacy trial — so read it as a tolerability signal, not proof of superiority — but it points the same direction.

The practical reframe matters: the most common reason a retinoid “doesn’t work” for someone is not that the molecule failed, but that the irritation made them quit before it could. An ingredient that lands on a similar collagen signal with a gentler ride isn’t a weaker option for those people — it may be the one they can actually stay on. Consistency is most of what makes any topical active work.

Where it sits next to retinol and tretinoin

It is worth being precise about the comparison, because the marketing blurs three different things into one. Retinol is a relatively gentle over-the-counter retinoid that your skin must convert through two steps into retinoic acid before it does anything. Tretinoin is retinoic acid itself — the prescription-strength, most-studied gold standard, with decades of trial data for photoaging that bakuchiol cannot remotely match in depth. Bakuchiol, again, is none of these chemically; it is a separate compound that happens to act a bit like the first one.

So the fair statement is narrow and specific: bakuchiol has been shown comparable to retinol — the gentle OTC tier — in one good trial, not comparable to tretinoin, the clinical benchmark. If your goal is maximal, evidence-backed remodeling and you can tolerate it, tretinoin remains the most-proven choice; our deep read on tretinoin and skin aging lays out why the depth of its trial record still sets the standard.⁶ Bakuchiol’s pitch is not “stronger than the gold standard” — it is “retinol-tier results, retinoid-style mechanism, less irritation,” which is a real and useful niche without being a giant-killer.

Where it fits: a tiered view

It helps to place bakuchiol honestly on a spectrum of who it is actually for, alongside the rest of the skin and aging toolkit.

Foundational — the non-negotiables first. No active competes with daily broad-spectrum sunscreen, which has the strongest evidence of anything in photoaging prevention, plus not smoking and adequate sleep. Bakuchiol is a remodeling nudge layered on top of prevention, not a substitute for it. If sunscreen isn’t daily, that is the higher-yield lever, every time.

Research-curious — the gentle on-ramp. If you want retinoid-style benefits but retinol’s scaling and stinging have repeatedly driven you off, bakuchiol is a low-irritation, plausibly retinol-comparable option with one good human trial behind it.¹ It is also a reasonable choice for sensitive, rosacea-prone, or eczema-prone skin that tends to flare on retinoids.⁵ Pairs naturally with a barrier-supporting active like niacinamide.

Experimental — betting it beats the gold standard. Treating bakuchiol as a full, evidence-equal replacement for tretinoin, or expecting it to out-perform a prescription retinoid, runs ahead of what the data support. The evidence base is small and short; it has not been tested at tretinoin’s depth or duration. As a substitute for the most-proven option in someone who tolerates that option, it is the weakest-supported use.

Grey areas: “natural” is not “risk-free”

The plant it comes from carries a real caution. Psoralea corylifolia — babchi — is naturally rich in psoralens, the same class of photosensitizing compounds used deliberately in medical light therapy precisely because they make skin react to UV. Raw babchi oil is not the same thing as cosmetic bakuchiol: commercial bakuchiol is a purified, isolated compound, and the bakuchiol molecule itself is not a psoralen. But this is exactly why “it’s natural, so it’s safe” is the wrong instinct. Botanical origin tells you nothing about safety; purity and dose do. Unstandardized “babchi oil” sold as a DIY bakuchiol is a meaningfully different and riskier product than a formulated cosmetic.

It can still irritate. “Better tolerated than retinol” is not “cannot cause a reaction.” The literature includes at least one documented case of contact dermatitis from a bakuchiol product, and any topical active can provoke irritation or allergy in a given individual.² Gentle on average is not gentle for everyone.

The product is not the molecule. “Bakuchiol” on a label tells you an ingredient is present, not at what concentration, in what vehicle, or alongside what else. The trial results were generated at a defined 0.5% in a specific formulation;¹ a serum with a trace of bakuchiol near the bottom of its ingredient list is not the same product and should not be expected to do the same thing. Concentration and formulation are doing real work, and they vary enormously across the market.

What we don’t know yet

No large, long, independent head-to-head. The entire efficacy case rests heavily on one 44-person, 12-week trial.¹ There is no large multi-site randomized comparison, no long-duration study of bakuchiol used for a year or more, and limited data on durability of effect once you stop. The single most useful study that does not yet exist is a big, independent, long-run bakuchiol-versus-retinoid trial.

It grades below the gold standard for a reason. A 2024 systematic review of antiaging cosmeceuticals graded the evidence formally and the gap is instructive: retinol and vitamin C earned the top recommendation grade, while bakuchiol earned a lower grade reflecting its thinner, smaller evidence base.⁴ That is not a verdict that bakuchiol fails — it is a measurement of how much less it has been studied. Grades track quantity and rigor of evidence, and bakuchiol simply has less of it.

Population gaps. The trials skew toward short timeframes and modest, fairly narrow samples. Data are thin in darker skin types across the full range of pigmentary outcomes, in pregnancy (where retinoids are contraindicated and a “safe alternative” claim would be especially consequential and is not established), and in long-term continuous use. “Promising in the studied populations” is the accurate ceiling.

What this article is not saying

This is not “bakuchiol is hype.” Within its limits, the human data are positive, the mechanism is genuinely retinol-like at the level of gene expression, and for the right person — especially anyone whose skin revolts against retinoids — it is one of the better-supported gentle actives available. Dismissing it outright is as wrong as overselling it.

This is not “bakuchiol beats retinol, let alone tretinoin.” It has been shown comparable to retinol in a single small trial, and it has not been tested against the prescription gold standard at anything like the same depth. The honest claim is “retinol-tier results with less irritation, on a thin evidence base” — not “a stronger, safer everything.”

And this is not a recommendation to swap your routine. It is a map of what the trials show and where they stop, so your expectations — and your shelf — can be honest ones. If you have a specific skin condition, a pigment disorder, or you’re pregnant, the relevant move is a clinician conversation, not a label swap.