TL;DR: Retinization is a 12 to 16 week neurological and structural reset, not a side effect to push through. Weeks 1-2: receptor binding and irritation. Weeks 3-6: peeling and the purge window. Weeks 7-12: pigment normalisation and texture changes. Weeks 13-24: actual collagen restructuring. Most people quit at week 3 when the literature predicts they would have quit.
The Brisbane reader I wrote about in the retinol breakup piece had a sister who messaged me a month later. She had started tretinoin 0.025% in late January after watching her sister recover from over-use. By mid-February she was peeling on her forehead, breaking out in places she had never broken out before, and her skin felt like sandpaper. She wanted to know if she should stop. Her sister had told her to stop. Her dermatologist had told her to push through.
I told her the dermatologist was right, but for reasons the dermatologist had probably not explained. What she was experiencing was not failure or sensitivity. It was retinization, which is a predictable 12 to 16 week process with surprisingly specific weekly markers, and most of the people who quit retinoids quit during the worst of it because nobody told them where in the process they were.
This is the piece I wish I had read when I started tretinoin in 2017 and quit in week 4 because I thought my skin had broken. I want to walk through what is actually happening, week by week, what the literature says is normal, and what the markers are that mean something is wrong and you should pull back.
What retinization actually means
The word “retinization” was coined in the 1990s by clinicians watching what happened when patients first started tretinoin or its cousins. The skin does not just react to a retinoid. It restructures around it. Retinoic acid binds to RAR-alpha, beta, and gamma receptors in the nucleus and changes the transcription of around 500 genes. Around half of those changes are immediate (cell cycle, inflammation). The other half are slow (collagen, glycosaminoglycans, melanocyte signalling, sebocyte differentiation). The visible drama of weeks 1 through 4 is the immediate transcription wave. The actual reason you started using it is the slow wave that arrives in months 3 through 6.
Mukherjee et al. (2006, PMID: 18046911) and the original Kligman tretinoin papers (1986, PMID: 3771853) both lay this out, although they use different vocabulary. The point is that the timeline is not arbitrary. The receptor pharmacology dictates it. So when someone tells you “your skin will adjust”, they are not handwaving. They are describing a specific molecular adaptation that takes about 12 weeks to complete and 24 weeks to mature.
I am going to use tretinoin 0.025% applied 3-4 nights a week as the reference protocol because it is what most of the literature uses. If you are on retinol, retinaldehyde, or adapalene, shift everything by roughly 2 to 4 weeks (slower for retinol, similar for adapalene, slightly faster for retinaldehyde at the top of its dose range).
Week 1: nothing visible, everything happening
For most people, the first 5 to 7 nights of tretinoin look uneventful. Mild tightness after application. Maybe a faint warming. The temptation here is to think the product is not working and to apply more, more often, or on a heavier schedule. This is the single most common error I have seen and the single biggest predictor of quitting in week 3.
What is happening underneath: retinoic acid is binding to nuclear receptors and the transcription cascade is starting. Stratum corneum cohesion is beginning to loosen. Sebocyte differentiation is shifting. None of this is visible.
What is normal: nothing visible, possibly mild tightness, maybe a faint dryness on the cheeks the morning after.
What is not normal: stinging that lasts more than 20 minutes after application, immediate redness that does not resolve overnight, swelling. If you have any of those in week 1, you started too high or too often. Drop to twice a week or buffer with moisturiser sandwiched on either side.
Weeks 2 and 3: the peeling window
This is where most people quit. The stratum corneum cohesion loss from week 1 becomes visible. Skin starts to flake, particularly around the nose, the corners of the mouth, the chin, and sometimes the forehead. The flaking is fine and powdery in lighter cases and sheet-like in heavier ones. Texture feels rough. Makeup catches on it.
Kligman et al. described this in 1986 (PMID: 3771853) as part of the expected initial response. Weiss et al. 1988 (PMID: 3336177) noted that around 90 percent of patients in their tretinoin trial reported some peeling in the first three weeks, and around 35 percent reported it severely enough to consider stopping. The Weiss trial is also where the 24-week clinical improvement endpoint comes from, which is to say: the very paper that established tretinoin’s photoaging benefit also documented that more than a third of users wanted to quit during the early peeling phase.
What is normal: fine to moderate flaking lasting 5 to 14 days, mild stinging on application, slight pinkness in the morning that resolves by noon, makeup that does not sit well, lipstick bleeding around the mouth corners.
What is not normal: open cracks, sheet-like peeling that lasts more than three weeks, weeping, persistent burning, eyelid swelling, eczema-pattern itch. Those mean you have crossed from retinization into actual barrier damage and you need to stop for 7 to 14 days and restart at a lower frequency.
The cheek and jawline areas peel less than the chin, nose, and mouth corners. This is partly sebum-driven (more sebum buffers more) and partly anatomical (skin around mobile features experiences more mechanical disruption). The maps do not lie. If only your cheeks are peeling and your chin is fine, you have applied unevenly.
Weeks 4 to 6: the purge, if it is going to happen
The purge is overhyped and underspecified. The actual mechanism is that retinoic acid accelerates the maturation of subclinical microcomedones, which means lesions that would have surfaced over the next 4 to 12 weeks all surface in a compressed 2 to 4 week window. Leyden, Stein-Gold, and Weiss (2017, PMID: 28585191) describe this as an expected consequence of follicular keratinisation normalisation. It is not a detox. It is a release of pre-existing lesions on a faster schedule.
What is normal: an increase in inflammatory acne lesions, particularly in areas you already get acne, peaking around weeks 4 to 6 and resolving by weeks 8 to 10. Some people get nothing. Some people get a moderate uptick. A small fraction get a significant flare.
What is not normal: lesions appearing in places you have never had acne before (rare but possible if the formulation contains a comedogenic excipient like isopropyl myristate; check the inactives), cystic nodules that were not present before treatment, perioral dermatitis-pattern bumps around the mouth (this often means you have inadvertently extended a moisturiser containing petrolatum into the perioral area combined with the retinoid, which is a recipe for it).
People who do not have active or subclinical acne often do not have a purge at all. The Mukherjee review covers this. If you are 38 and using tretinoin for photoaging and your skin was clear when you started, you will likely skip weeks 4 to 6 of the purge entirely and just have peeling and adjustment.
Weeks 7 to 12: the pigment phase
This is the phase nobody talks about. Once the peeling subsides and the purge resolves, skin enters a 4 to 6 week period where pigmentation patterns shift. Existing post-inflammatory hyperpigmentation accelerates its fade (good). New marks from the purge phase look angry for several weeks (bad, but transient). For Fitzpatrick III and above, this is when sunscreen discipline matters most. The retinoid has accelerated epidermal turnover, which means melanocytes are more sun-reactive and any UV exposure during weeks 7 to 12 will leave pigment that takes longer to fade than it would have otherwise.
What is normal: fading of old PIH, slight darkening of new PIH from purge lesions, occasional patches of mild dryness, returning makeup compatibility, skin starting to feel different (smoother, more even) by the end of week 12.
What is not normal: new sun-pattern pigmentation (the Mallar mask, mottling on the cheekbones, dark patches that follow no acne distribution). This means UV exposure has caught up with you and you need to tighten sunscreen.
Kang et al. (2001, PMID: 11735711) ran a 24-week head-to-head of tazarotene and tretinoin and noted that meaningful pigment improvement scores diverged from baseline at week 12 in both arms. Weiss et al. 1988 saw the same pattern. Week 12 is when the pigment changes start to be measurable. Before that, you are mostly seeing the noise of peeling and purge.
Weeks 13 to 24: collagen, the actual point
Here is the part the marketing copy on retinol skips. The reason retinoids treat photoaging is not that they exfoliate or that they fade pigment. It is that they cause type I and type III procollagen synthesis to increase by around 80 percent over baseline by week 12, sustained through week 24 and beyond. Talwar et al. (1995, PMID: 7714086) showed this with skin biopsy data. The collagen accumulating in the upper dermis is what causes the smoothing of fine lines that everyone is chasing.
The thing about collagen is that it accumulates linearly. The first 12 weeks build the infrastructure. The next 12 weeks deposit measurable matrix. The next 12 after that fine-tune it. People who quit at week 8 because they “did not see results” left before the result was scheduled to arrive.
What is normal: gradual smoothing of fine lines starting around week 16, slight plumping of texture, less visible pores (technically a perception effect, not anatomical), tightening around the eye and mouth contours.
What is not normal: lines getting deeper or skin getting thinner-looking. This usually means you are over-using and have crossed from collagen-up to inflammation-driven matrix metalloproteinase activation, which actually degrades collagen. If you are using tretinoin nightly and 0.05% or higher and your skin looks worse at month 4 than at month 1, you are over-dosing.
What “normal” looks like at month 6
After six months on tretinoin 0.025% three to four nights a week, the typical patient in the Weiss 1988 trial showed: 50-60% reduction in fine periorbital lines on physician scoring, 30-40% reduction in coarse mottling, modest reduction in tactile roughness. Not dramatic. Visible. The dramatic visual transformations on Reddit before-and-afters usually involve at least one of: lighting changes, makeup, weight changes, healing of incidental acne, or 12 to 24 months of use rather than 6.
What I changed in my routine
I stopped buffering with moisturiser when I started, because the buffering was the wrong intervention for my problem. I was applying too often, not too strongly. I dropped to twice a week and let the peeling phase happen at the predictable time, and then escalated to nightly over months 3 to 4. The total time to comfortable nightly use was about 14 weeks. Trying to compress that compressed the irritation but not the underlying receptor adaptation.
I started photographing the same three angles on the first of each month under the same light. The peeling phase distorted everything visually for 3 weeks. Without the monthly photos I would have convinced myself, multiple times, that my skin had got worse. The photos showed a slow, monotonic, slightly boring improvement.
I bought one of the cheap pH 5.0 to 5.5 cleansers and used it on retinoid mornings and nothing else, and I stopped using exfoliating acids entirely for the first 12 weeks. This was probably the single highest-value change.
FAQ
Should I buffer my retinoid with moisturiser?
Mostly, no. The buffering literature is thin and most clinical trials apply the retinoid to dry skin. Buffering can reduce irritation in week 1 but it also reduces efficacy proportionally, and you end up extending the retinization timeline rather than completing it. Better to lower the frequency, not dilute the dose.
Why does my skin feel rough between weeks 6 and 10 when the peeling has stopped?
The peeling-to-smooth transition is not instant. Stratum corneum cohesion is normalising at the new equilibrium and the corneocytes that surface during this period are slightly less organised than usual. By week 12 the new equilibrium establishes and the roughness resolves. If it has not by week 14, you are probably under-moisturising rather than over-treating.
Is the purge real or marketing?
The release-of-microcomedones mechanism is real and biologically plausible. The word “purge” is marketing. About 30 to 50 percent of acne-prone retinoid starters experience a measurable uptick in lesions between weeks 4 and 8, resolving by week 10. People who do not have active or subclinical acne usually skip the phenomenon entirely.
What if I miss a few weeks because of a holiday?
You will lose roughly half of the receptor adaptation you have built. Restarting after a 3-week gap means going back to 2 nights a week for 2 weeks before resuming your previous frequency. After a 6-week gap, restart from scratch. The receptor desensitisation that allowed you to tolerate nightly application reverses faster than it accumulated.
Should I switch retinoids if I am not seeing results at week 8?
Almost never. Week 8 is too early to assess. The single most common error after “quitting in week 3” is “switching at week 8 because I am impatient.” Give it 16 weeks at the same product and frequency before changing anything.
What I would tell my past self
Retinization is a 12 to 16 week neurological and structural reset, and most of the irritation you are reading about online is happening to people in weeks 2 and 3 who quit before week 5. Do not start during a season when your skin is already stressed (winter dryness, post-holiday travel, exam week). Do not start two new things at the same time. Do not start at 0.05% because the influencer you watched is on 0.1%. Start at 0.025% twice a week and add a third night only when the previous frequency has been comfortable for 14 consecutive days. Take photographs. Wait until week 16 to form an opinion about whether it is working. You will be wrong about that opinion if you form it earlier.
Related Elelaf tools
Sources
- Mukherjee S, Date A, Patravale V, Korting HC, Roeder A, Weindl G. Retinoids in the treatment of skin aging: an overview of clinical efficacy and safety. Clin Interv Aging. 2006;1(4):327-348. PMID: 18046911
- Kang S, Leyden JJ, Lowe NJ, et al. Tazarotene cream for the treatment of facial photodamage: a multicenter, investigator-masked, randomized, vehicle-controlled, parallel comparison of 0.01%, 0.025%, 0.05%, and 0.1% tazarotene creams with 0.05% tretinoin emollient cream applied once daily for 24 weeks. Arch Dermatol. 2001;137(12):1597-1604. PMID: 11735711
- Kligman AM, Grove GL, Hirose R, Leyden JJ. Topical tretinoin for photoaged skin. J Am Acad Dermatol. 1986;15(4 Pt 2):836-859. PMID: 3771853
- Weiss JS, Ellis CN, Headington JT, Tincoff T, Hamilton TA, Voorhees JJ. Topical tretinoin improves photoaged skin. A double-blind vehicle-controlled study. JAMA. 1988;259(4):527-532. PMID: 3336177
- Leyden J, Stein-Gold L, Weiss J. Why Topical Retinoids Are Mainstay of Therapy for Acne. Dermatol Ther (Heidelb). 2017;7(3):293-304. PMID: 28585191