Why Your Retinol Might Not Be Working

Read time: 13 minutes

It has been called a miracle, a myth, a “gold standard” and a pregnancy hazard, sometimes in the same Instagram carousel. Half the internet swears it erased their wrinkles. The other half is convinced it thinned their skin into tissue paper or that one accidental pump will harm a pregnancy.

The truth is more interesting than either camp, and it lives in a stack of clinical trials that almost nobody reads for fun.

So I read them for you.

Here’s what holds up, what’s wishful thinking, and what’s just very good packaging.

First, “retinol” is not one thing

This is the misunderstanding that powers much of the confusion, so it is worth clearing up first.

Retinoid is the family name. It’s the umbrella term that includes vitamin A and a range of natural and synthetic chemical relatives.

Retinol is just one member of that family.

The difference that separates the relatives comes down to how many conversion steps are required before it gets to the form that produces the main retinoid effects in the skin.

Your skin can only respond to one form of vitamin A: retinoic acid. Everything else is a precursor that has to be converted, step by step, before it does anything. Think of it as a relay race where the baton is the active molecule, and only the runner crossing the finish line counts.

• Retinyl esters, such as retinyl palmitate, sit furthest back in the conversion chain. They require several steps and have the weakest evidence for visibly improving ageing skin.

• Retinol requires two conversions before becoming retinoic acid. It is widely available without a prescription and is generally less irritating than tretinoin, but its performance varies considerably between products.

• Retinaldehyde, usually shortened to retinal, requires one conversion. This makes it an appealing middle ground between retinol and prescription treatments, although that does not automatically mean every retinal formula will outperform every retinol product.

• Tretinoin is retinoic acid. It does not require conversion, which helps explain both its effectiveness and its greater potential to irritate.

There are synthetic retinoids too, including adapalene, tazarotene and trifarotene. These do not simply follow the same conversion ladder. They are designed to interact selectively with particular retinoic acid receptors.

The neat hierarchy is useful, but it hides something important.

Potency on paper is not necessarily potency in the bottle.

A well-formulated, stable retinol can easily outperform a carelessly stored retinal. Which brings us to one of the least discussed problems in the category.

Why “0.5% retinol” means almost nothing on its own

Retinol is a diva. It degrades in light, heat, and oxygen the way a cut avocado browns on the counter.

Two products both labelled “0.5% retinol” may therefore behave very differently. The cream or serum base, packaging, storage conditions and delivery technology can all influence how much remains active.

Concentration also does not tell you how much of the ingredient penetrates the skin, how quickly it is released or how well the formula protects it from degradation.

It gets murkier. Most clinical trials don’t test pure retinol. They test a finished commercial cream loaded with moisturisers, antioxidants, and other actives. So when the skin looks better at week 12, it’s genuinely hard to say how much credit goes to the retinol and how much to everything else in the tube.

This is the single most useful point to remember about the entire category:

Retinoids work. That does not mean every product with “retinol” on the label has been proven to work equally well.

Hold onto that. It quietly debunks about 70% of the marketing you’ll see this year.

What retinoids actually do (it’s not exfoliation)

Retinoids are often described as exfoliants, but that explanation misses most of what makes them interesting.

They do influence the way cells are shed and replaced, but their deeper effects occur through changes in gene expression.

Once retinoic acid gets inside a skin cell, it binds to receptors in the nucleus and changes which genes get switched on. It’s less like a scrub and more like a manager walking the floor and reassigning the workers.

In sun-damaged skin, UV light has spent years activating enzymes called matrix metalloproteinases or MMPs, the demolition crew that chews through your collagen and leaves the dermis disorganised. Retinoids do two things at once: they tell the demolition crew to stand down, and they nudge your fibroblasts (the cells that build collagen) to get back to work laying down fresh material.

That’s the whole magic. Less breakdown, more rebuilding. Over months, that shows up as smoother texture, more even tone, and softened fine lines.

Over time, retinoids can improve:

• fine lines

• rough texture

• mottled pigmentation

• some sun spots

• dullness

• acne and blocked pores

• shallow wrinkles associated with sun damage

But expectations matter.

A retinoid is a good editor, not a ghostwriter.

It can improve the structure already there. It cannot replace lost facial fat, tighten significantly loosened ligaments or erase deep folds caused by facial movement and volume loss.

A cream cannot do the job of a facelift, filler or resurfacing procedure.

That does not make retinoids ineffective. It means their effects are gradual and biological rather than transformative.

The evidence, ranked honestly

Here’s where the two big recent reviews earn their keep. A 2024 systematic review of 25 comparative studies and a 2025 network meta-analysis pooling 23 trials and nearly 4,000 people both took a hard look at the whole field. The verdict is refreshingly un-hyped. Together, they show that retinoids can improve visible photoageing, but they also expose how uneven the evidence remains.

Tretinoin: the benchmark. Prescription tretinoin remains the gold standard, full stop. Decades of randomised trials show it improves fine and coarse wrinkles, pigmentation, and roughness, with actual histological proof of skin remodelling underneath.

A 2022 systematic review of randomised controlled trials found consistent improvements in multiple signs of photoageing, with benefits continuing during treatment periods extending as long as 24 months. The 2024 review kept it as first-line therapy not because alternatives don’t work, but because the evidence for them simply isn’t as strong yet.

Tretinoin also illustrates why stronger is not always better.

A double-blind comparison of 0.025% and 0.1% tretinoin found similar improvements in photoageing, while the stronger concentration produced more irritation.

The best dose is often the one a person can continue using, not the strongest one available.

Retinol: biologically credible, commercially chaotic. Lab studies show retinol stimulates the right collagen pathways, and the 2025 meta-analysis found real effects on several ageing measures.

But the 2024 systematic review was less reassuring about the quality of the supporting studies. Many were small, commercially connected or tested combination products that made the effect of retinol difficult to isolate.

The fairest conclusion:

Retinol probably can improve sun-damaged skin, but you cannot treat every bottle on a pharmacy shelf as a miniature version of prescription tretinoin.

Retinaldehyde: the sensible compromise. Retinal sits one conversion step closer to retinoic acid than retinol.

Some clinical studies suggest it can improve fine lines and texture while being more tolerable than tretinoin.

The 2024 review considered retinaldehyde and related precursor technologies reasonable second-line options for people who cannot tolerate tretinoin.

But the evidence is not yet strong enough to declare retinal universally superior to retinol or equivalent to prescription treatment.

It is a credible option, not a guaranteed shortcut.

Adapalene: the quiet overachiever. Best known for acne, more photostable and better tolerated than old-school tretinoin. One study found adapalene 0.3% gel comparable to tretinoin 0.05% cream for ageing. Encouraging, though the anti-ageing evidence base is still small.

However, the anti-ageing evidence for adapalene remains much smaller than the evidence supporting its use for acne or the evidence supporting tretinoin for photoageing.

Tazarotene: powerful, prickly. Strong on coarse wrinkles, but it had the worst tolerability of the bunch. Which raises a recurring theme worth pausing on.

A potent product that makes your skin so irritated you quit after three weeks will, in the end, do less for you than a gentle one you actually keep using.

The irritation myth

Somewhere along the way, the skincare internet absorbed a piece of gym-bro logic and applied it to faces: no pain, no gain.

If your skin is red, peeling, and stinging, the thinking goes, the retinol must be working.

It isn’t. Your face is not a CrossFit class.

Redness, burning, flaking, and tightness are signs of irritation and a disrupted skin barrier, not signs of collagen being built. Irritation is not required for the treatment to work.

Persistent inflammation may also worsen pigmentation, particularly in skin that is prone to post-inflammatory hyperpigmentation.

The early adjustment period sometimes called “retinisation” is real. Mild dryness, tightness or flaking can occur when a retinoid is introduced and may settle as the skin adapts. But the term is often used to justify irritation that has moved well beyond a reasonable adjustment period.

Mild, temporary flaking is one thing. Painful burning, cracking, swelling, severe redness or worsening eczema are signals to reduce the frequency, stop temporarily or seek professional advice.

Tolerance also varies between people.

Eczema, rosacea, an already impaired skin barrier and a tendency toward post-inflammatory pigmentation can all make an aggressive introduction less forgiving.

There is no trophy for applying a retinoid every night if your skin can only tolerate it twice a week.

One way to reduce irritation is the “sandwich” method: moisturiser, followed by a pea-sized amount of retinoid, followed by another layer of moisturiser.

This may reduce penetration slightly, but a routine that is tolerated and continued is likely to outperform one that is theoretically perfect and quickly abandoned.

“But what about the pill version?”

This is where things go sideways, because two very different “oral vitamin A” stories get blended into one.

Oral isotretinoin is a powerful prescription treatment for severe, scarring or treatment-resistant acne. It reduces the size and activity of sebaceous glands, changes follicular keratinisation and has anti-inflammatory effects.

Because it can dramatically reduce acne and oiliness, skin may look smoother and pores may appear less visible after treatment. That does not make it a general anti-ageing medication.

Studies of low-dose oral isotretinoin for photoageing have produced mixed and generally unconvincing results. A 2023 review of oral isotretinoin’s dermatological uses concluded that evidence for treating photoageing was insufficient once the drug’s risks were considered.

Oral isotretinoin is an important medical treatment for the right patient.

It is not a longevity supplement or a wrinkle shortcut.

Oral vitamin A supplements are the other half of the confusion, and the logic sounds seductive: if topical vitamin A helps my skin, surely swallowing it helps from the inside?

No. And this one can actually hurt you.

Vitamin A is fat-soluble. Excess preformed vitamin A is stored rather than simply excreted. High intakes can cause liver injury, headaches, bone effects, hair loss, dry skin and birth defects.

The US National Institutes of Health sets the adult upper intake level for preformed vitamin A at 3,000 micrograms of retinol activity equivalents per day, equivalent to approximately 10,000 IU of retinol.

That is an upper safety limit, not a cosmetic target.

There is no strong evidence supporting high-dose vitamin A supplementation as a routine anti-ageing strategy.

High-dose vitamin A should not become the next collagen gummy.

The pregnancy question

This is the part that sends people into a 2 a.m. panic spiral.

The fear is real, and it’s earned, but it didn’t start with face cream. It started with the pill.

Retinoic acid isn’t just a skincare molecule; in a developing embryo it’s a master signalling switch that helps lay out the face, ears, heart, brain, and more. Flood that delicate system with extra retinoid at the wrong moment and development can go badly wrong. The result is a recognised pattern of serious malformations.

Oral isotretinoin is a confirmed, potent human teratogen.

This is not a theoretical warning or an excess of regulatory caution.

The UK Teratology Information Service reports substantial risks of congenital malformation and miscarriage following pregnancy exposure.

Estimates vary because studies count pregnancies and outcomes differently, but reported major-malformation risks have commonly fallen between approximately 5% and 28% among prospectively followed live births. Other sources report still higher figures when broader groups of exposed pregnancies are considered.

Miscarriage risk may also be increased.

No dose or duration has been established as risk-free, which is why oral isotretinoin is contraindicated throughout pregnancy.

It is also why treatment is accompanied by strict pregnancy-prevention requirements, including pregnancy testing and reliable contraception before, during and for a period after treatment.

An oral isotretinoin exposure during pregnancy requires urgent medical and specialist assessment.

Because the family contains a proven teratogen, the entire family gets treated as guilty by association, including the cream.

Topical retinoids: a different risk question

Topical retinoids create much lower systemic exposure.

When used as directed on intact skin, only a small amount is expected to enter the bloodstream. Absorption may increase when a product is applied excessively, over a large area or onto damaged or inflamed skin.

Human observational data have so far been broadly reassuring.

A systematic review and meta-analysis of first-trimester topical-retinoid exposure did not identify a significant increase in major congenital malformations, miscarriage, stillbirth, low birth weight or premature birth.

Other cohort studies have reached similar conclusions.

A small number of concerning case reports do exist, including reports of abnormalities resembling those associated with systemic retinoids. Case reports cannot establish that the topical product caused the outcome, but they do contribute to the remaining uncertainty.

So why do regulators still advise avoiding topical retinoids during pregnancy?

Because reassuring observational evidence is not the same as proof of zero risk.

A randomised trial deliberately exposing pregnant women would be unethical. Existing studies cannot completely exclude very rare effects. The drug family contains a powerful known teratogen, and the benefit being sought is usually cosmetic rather than medically essential.

The European Medicines Agency states that systemic absorption from topical retinoids is negligible and fetal harm is unlikely, but it still recommends that topical retinoids not be used during pregnancy or when planning pregnancy as a precaution.

Both facts can be true at once:

• Accidental topical exposure appears to carry a very low risk.

• Continuing a cosmetic retinoid during pregnancy is not recommended.

(None of this is personal medical advice. If you’re pregnant, planning, or breastfeeding, your own clinician gets the final word.)

What if you used retinol before knowing you were pregnant?

Using a retinol serum or topical tretinoin before discovering a pregnancy is not equivalent to taking oral isotretinoin.

MotherToBaby’s guidance on topical tretinoin notes that studies of correct topical use have not identified an increased overall rate of birth defects, although avoidance during pregnancy is still advised.

The practical response is to stop the product, record exactly what was used and discuss the exposure with the maternity care provider.

It is reasonable to seek individual advice.

Accidental topical exposure usually calls for calm assessment and reassurance. Oral isotretinoin exposure calls for urgent specialist advice.

Those are not the same situation.

What about breastfeeding?

Oral isotretinoin is not recommended during breastfeeding.

Topical tretinoin has not been studied extensively in breastfeeding, although the amount expected to enter breast milk is likely to be low because systemic absorption is minimal.

It should not be applied to the breast, nipple or any area that may contact an infant’s skin or mouth.

Because recommendations can vary depending on the exact product, dose and area being treated, this is best discussed with the prescribing clinician.

If retinol’s off the table, what works instead?

Good news: pregnancy doesn’t mean surrendering your skin for nine months.

The standout alternative is bakuchiol, a plant-derived compound that’s chemically nothing like vitamin A. It doesn’t touch the retinoid receptors or convert to retinoic acid.

Yet it seems to nudge similar collagen pathways.

In a 2018 randomised trial, 0.5% bakuchiol went head-to-head with 0.5% retinol and produced comparable improvements in lines and pigmentation, with less stinging and flaking.

The honest caveat: there are no large safety studies in pregnant women specifically (there can’t be, ethically), and it hasn’t been formally rated for pregnancy. Therefore, it is more accurate to describe it as a promising retinoid-free option than as a product that has been definitively proven safe during pregnancy.

Two other well-established options: azelaic acid, which tackles pigmentation and breakouts and has a reassuring pregnancy track record, and glycolic acid at low strengths for gentle resurfacing.

So what should you actually do?

If you take one practical thing from all of this, make it this:

Sunscreen does more for preventing visible ageing than any retinoid does for reversing it.

It’s genuinely a little absurd to run a nightly retinoid to repair collagen while spending your days undoing that exact work in the sun. The retinoid is bailing water; sunscreen patches the hole.

A sane, evidence-backed routine looks boring, which is the point:

• Broad-spectrum sunscreen, every day, non-negotiable

• A gentle cleanser

• A barrier-supporting moisturiser

• One retinoid, introduced slowly

• Not a teetering stack of acids, scrubs, and exfoliants fighting each other

For the retinoid itself: a pea-sized amount for the whole face. Start two or three nights a week and build up as your skin allows. Sandwich it with moisturiser if it bites. And judge results over months, not days.

Collagen doesn’t work on a content-creator’s posting schedule.

A tolerable routine you actually keep beats an aggressive one you rage-quit. Every single time.

Coming Up Next: The Practical Guide

One of the strangest things about retinoids is that two products with the same percentage can produce completely different results.

Next we’ll look at how to choose a retinoid, what actually matters on the label, and which marketing claims deserve a healthy dose of scepticism.

The Clarity

The biggest misconception in skincare isn’t that retinoids don’t work.

It’s assuming every product that says “retinol” works equally well.

They don’t.

Understanding that one distinction will probably save you more money than any skincare tip you’ll hear this year.