The circadian system is one of the few areas in sleep science where the mechanism is well-mapped and the interventions are clean. Light hitting the retina at the wrong time delays or advances the internal clock. Melatonin, dosed at the right time and at the right amount, does the same thing chemically. If you know your target sleep window and the direction you need to shift, you can move a human circadian rhythm by roughly an hour a day. That is the ceiling. Anyone selling you a faster reset is selling you something else.
This piece is about three separate problems, because they need three separate protocols. Jet lag is an acute mismatch after travel. Night shift work is a chronic mismatch by employment. Chronic phase delay — delayed sleep-wake phase disorder in the ICSD-3 — is a mismatch by biology. The tools overlap. The timing does not.
The one-hour-a-day ceiling
The human circadian pacemaker sits in the suprachiasmatic nucleus of the hypothalamus. Its intrinsic period runs slightly longer than 24 hours in most adults — Czeisler's 1999 landmark study in Science pinned the population mean at about 24.18 hours. It gets entrained to the actual 24-hour day mostly by light, and secondarily by activity, meals, and social cues.
The published shift rate under aggressive protocols is roughly 1 to 1.5 hours per day when you push in the correct direction, and closer to 0.5 hours per day when you resist the direction the clock naturally drifts. Eastman and Burgess have documented this repeatedly across two decades of shift-work and jet-lag trials. Do the math before you plan a trip. A New York-to-Tokyo shift is 14 hours. You are not fixing that in one day. You are managing it.
Light is the primary lever, not melatonin
The mistake most people make is treating melatonin as the sleep drug and light as a background variable. It is the other way around. Light exposure to the retina is the strongest zeitgeber the circadian system has, by roughly an order of magnitude. Melatonin, dosed correctly, is a useful adjunct that nudges the phase-response curve. Dosed incorrectly — usually too much, too late — it is a mild sedative that does nothing for your circadian phase.
The phase-response curve to light, mapped by Khalsa and colleagues in 2003, tells you everything about direction. Bright light in the hours before your body-clock nadir (roughly 2 to 4 AM in a normally-entrained adult) delays the clock. Bright light in the hours after the nadir advances it. The nadir moves when your schedule moves, which is why timing is the whole game.
Practical intensities matter. Melatonin suppression starts around 10 lux at the eye. A meaningful phase shift usually requires 2,500 to 10,000 lux for 30 to 60 minutes, or bright outdoor light — which is 10,000 to 100,000 lux on a sunny day and 1,000 to 10,000 lux on an overcast one. A lamp on your desk is not doing this. A 10,000 lux light-therapy box at 12 to 16 inches from the face is.
Protocol 1: Jet lag
Jet lag follows a rule with almost no exceptions. Eastward travel is harder than westward travel of the same distance. Going east requires you to advance your clock — go to bed earlier, wake earlier — and that runs against the intrinsic 24.18-hour drift. Going west requires you to delay your clock, which the drift favors.
Westward travel (delay)
You want later bedtimes and later wake times at destination. Start about 3 days before departure.
- Push bedtime and wake time later by roughly 1 hour per day toward destination time.
- Get bright light in the evening at the pre-departure end of your day — outdoors after 6 PM, or a 10,000 lux box for 30 minutes.
- Avoid morning light at the pre-departure end. Sunglasses on the walk to the car if you must go out.
- On arrival, get afternoon and early-evening outdoor light. Try not to nap. Eat at destination meal times.
Eastward travel (advance)
This is the harder direction. Eastman's 2005 protocol is the cleanest published playbook.
- Push bedtime and wake time earlier by 1 hour per day for 3 days before departure.
- Bright morning light for 30 to 60 minutes on waking — outdoors is best, 10,000 lux box otherwise.
- Avoid evening light. Sunglasses on the commute home if the sun is still up.
- Low-dose melatonin, 0.3 to 0.5 mg, taken roughly 5 hours before target destination bedtime. This is the phase-advancing dose, not the sedating dose.
- On arrival, get bright morning light immediately. Avoid outdoor light before local dawn. Continue low-dose melatonin at target bedtime minus 5 hours for the first 3 to 4 nights.
You will not be fully adjusted on day one. You are aiming for functional — able to work, drive, and not fall asleep at 3 PM — not perfect. Full alignment takes 4 to 7 days for a large eastward shift.
Protocol 2: Night shift work
Night shift is not jet lag repeated. Jet lag ends when you get home. Night shift is chronic circadian misalignment, and roughly 15 to 20 percent of workers who do it never fully adapt on any measurable schedule. Shift Work Disorder is a real ICSD-3 diagnosis. The best you can do is protect the sleep you get and defend the circadian phase you build during your work stretch.
The anchor sleep protocol
Fixed permanent night shift is easier to adapt to than rotating shifts, but most night shift workers do not have that luxury. The anchor sleep approach — documented in Folkard's work through the 1990s and 2000s — keeps a consistent 4-hour block of sleep at the same clock time every day, whether working or not, and layers additional sleep around it. On workdays, the anchor sits inside the daytime primary sleep. On days off, it stays put and the rest of the sleep gets added around it.
This is unglamorous and it works better than trying to flip fully back to a nights-off schedule every weekend. Flipping every few days is the worst circadian pattern a human can adopt short of no sleep at all.
Light and dark on the shift
- Bright light exposure early in the shift — 2,500 to 10,000 lux for the first 3 to 4 hours — improves alertness and helps delay the circadian phase toward a night-oriented schedule. Czeisler's phototherapy work in the late 1990s demonstrated this in controlled lab conditions and has been replicated in field studies since.
- Sunglasses on the commute home. Wraparound dark sunglasses, or amber blue-blockers if you are aggressive about it. Morning sunlight after a night shift is a phase-advancing signal that fights the phase delay you spent all night trying to build.
- Blackout at home. This matters more for shift workers than anyone else. A pitch-black bedroom during daytime sleep is not optional.
Strategic caffeine
Caffeine has a half-life of roughly 5 to 6 hours in most adults, longer in slow metabolizers and pregnant women. The functional rule for night shift: caffeine in the first half of the shift, none in the second. A 200 mg dose at the start of an 11 PM shift is largely cleared by 8 AM when you need to sleep. The same 200 mg at 4 AM is still meaningfully active at noon. Schweitzer and colleagues have shown that strategically timed low-to-moderate caffeine improves shift alertness without wrecking post-shift sleep. Coffee at hour 7 of a 12-hour shift wrecks post-shift sleep almost every time.
Protocol 3: Chronic phase delay
Delayed sleep-wake phase disorder is not "being a night owl." It is a persistent, biological misalignment between the internal circadian phase and the desired schedule, with a body-clock timing that runs 2 or more hours later than typical. It is common in adolescence — a real developmental phase shift documented by Carskadon since the 1990s — and it persists into adulthood in a meaningful minority. Roenneberg's work on chronotype distribution across large European populations makes it clear that late chronotypes are not a moral failing. They are a distribution.
The problem is when the biological phase does not fit the required schedule. A late chronotype who works from home and starts at 11 AM has no problem. A late chronotype who has a 7 AM job has DSWPD-shaped misery.
The morning-light protocol
- Bright light at wake time — 10,000 lux for 30 minutes, or the equivalent outdoors — every morning, including weekends. This is the primary intervention.
- Progressively earlier wake time — advance by 15 to 30 minutes every 3 to 4 days until the target wake time is reached. Fast enough to make progress, slow enough not to collapse.
- Dim light in the evening — under 30 lux at eye level for the 2 to 3 hours before bed. Real dim, not "night mode." That means low-wattage warm bulbs, not overhead lighting, and no bright screens.
- Low-dose melatonin (0.3 to 0.5 mg) roughly 5 hours before target bedtime. This is the phase-advancing use of melatonin — timed against dim-light melatonin onset, not against subjective sleepiness. Consumer bottles at 3 to 10 mg are 10 to 30 times the useful dose for this purpose. The extra milligrams do nothing productive; they may worsen next-day grogginess.
- Hold the schedule. Weekend sleep-ins destroy the progress. This is the intervention people fail on most.
Realistic expectations. Sustained adherence for 4 to 8 weeks usually produces a phase advance of 1 to 3 hours. Some late chronotypes never fully advance to typical morning-oriented schedules, which is a treatment consideration, not a failure of the protocol. Chronotherapy — a controlled progressive delay around the clock — is an older approach with real risks of overshoot and is not first-line anymore.
Melatonin, honestly
Consumer melatonin is one of the most misused supplements on the market. Two facts matter.
The dose most bottles ship at is wrong for phase-shifting. Brzezinski's 2005 meta-analysis and Burgess and Eastman's protocols converge on 0.3 to 0.5 mg as the effective phase-shifting dose. The 3, 5, and 10 mg pills flooding the market work as mild sedatives — a different thing.
The timing is the whole intervention. Melatonin 5 hours before target bedtime is a phase-advancing dose. The same dose taken 30 minutes before bed does approximately nothing to your circadian phase. It might make you drowsy — separate mechanism.
Tracking the reset
The one honest way to know your protocol is working is to measure sleep time, latency, and consistency across the 4- to 8-week window it takes to shift a rhythm. A wearable, an under-mattress sensor, or a paper sleep diary all work — you need the trend, not the millisecond precision.
The mat is not a diagnostic device. It will not tell you your dim-light melatonin onset. It will tell you, unambiguously, whether your average sleep midpoint moved earlier over the course of a 6-week phase-advance protocol, which is the outcome that matters. That is what "circadian reset" looks like in real data — a slow drift of the middle of your sleep block, week over week, in the direction you set.
What does not work
A few things marketed at circadian problems have essentially no evidence base.
- "Jet lag pills" that are not low-dose melatonin. B-vitamin and adaptogen blends have no credible phase-shift data.
- Grounding mats and earthing sheets. No evidence that ground contact shifts circadian phase.
- High-dose melatonin at bedtime as a shift-work tool. Sedating, not phase-shifting.
- Blue-light glasses used for a few minutes before bed. Real melatonin-onset advance requires amber or red lenses for 2 to 3 hours pre-bed.
The honest bottom line
Circadian rhythm reset is one of the few sleep interventions where the science is unambiguous. Light in the right window shifts the clock. Melatonin at the right dose and timing shifts it further. Nothing shifts it faster than about an hour a day. The direction depends on whether you are travelling east or west, working nights or days, and whether your biological clock runs early or late relative to the schedule you have to keep.
Pick the protocol that matches your problem. Run it consistently for 3 to 8 weeks depending on the size of the shift. Measure your sleep midpoint every week and watch it move. That is what a reset is.
