Nothing in your body is counting down the stall#
A flat scale that has held for under three weeks is usually not a plateau, and it clears on its own within days once fluid and gut contents settle. A genuine stall — a seven-day average flat for a month, with your waist flat alongside it — has no duration you can look up, because there is no timer. It persists until the gap between what you eat and what you burn reopens. Nothing in your physiology is scheduled to reopen it for you, and the best record of what actually happens next says the common sequel is not resumption.
What the calendar can tell you is whether you have flattened early or on schedule. Time-to-nadir has been measured for every major way of losing weight, and it ranges from about six months to about two years depending on how the deficit is created. That timetable is a genuinely useful reference point, and it is the closest thing to an answer this question has. Below: what follows a plateau in 667 people's weight records, when each intervention flattens, what waiting looks like when somebody measures it for eight years, and which of the three things called a plateau actually has a duration.
The only flat spell with a real clock is the fluid one#
One category does resolve by itself on a predictable schedule, and it is the category most people are in when they start worrying. Day-to-day body mass carries a spread larger than a week of good fat loss, so a week or two of flat readings is the expected appearance of a diet that is working normally — the arithmetic is worked out in plateau versus normal fluctuation. Water bound to glycogen, a salt-heavy weekend, a new training block and a hormonal phase all move the number by amounts that clear within days, because your kidneys clear them. That is why the standard advice is three weeks of flat weekly averages before you call anything: not caution for its own sake, but the point at which the fluid explanation runs out of runway. The sudden multi-day drop that follows one of these stalls has its own folklore and its own surprisingly specific physiology.
Everything from here assumes you are past that window.
What follows a plateau, in 667 people's weight records#
The most direct evidence on what a stall does next comes from an unglamorous source: the weigh-in logs of an online weight-management program. Researchers took 667 participants and 69,363 individual weight records — a mean loss of 7.9 ± 5.1% of bodyweight over 133 ± 69 days — and fitted each person's curve to find how many distinct shapes the data contained. Four emerged: a steady decrease over time (30.6%), a decrease to a plateau followed by further decline (15.8%), a decrease to a plateau followed by an increase (46.9%), and no substantial decrease at all (6.7%)1.
Read the middle two rows against each other, because that comparison is the answer to the question in the title. Among the roughly 63% of people whose curve contained a plateau at all, close to three in four went on to regain and about one in four resumed losing — that ratio is my arithmetic on their two percentages, not a comparison the paper draws. A plateau, in the largest set of real weigh-ins anyone has published on this, is not a pause. It is a fork, and the more travelled branch points up.
The caveats matter and they cut in one direction. These are self-reported weights from people who chose to use a commercial program, the trajectory classes are curve-fits rather than clinical events, and people who quit logging vanish from the record — which almost certainly makes the regain branch larger than it looks, not smaller. What the study cannot tell you is what would have happened to any individual who changed something, because nobody was randomized to anything. It describes the base rate for doing what most people do at a stall, which is wait.
The timetable you can check yourself against#
Here is where the calendar earns its place. The point at which weight loss flattens has been measured across every major intervention, and the numbers are far apart.
| How the deficit is created | When the curve flattens | What it does afterward |
|---|---|---|
| Behavioral diet and activity, with ongoing support | around year 1, at −8.5% | slow regain to −4.4% by year 4, then steady: −4.7% at year 8 |
| Semaglutide 2.4 mg | around week 60 | holds: −15.6% at week 52, −15.2% at week 104 |
| Bariatric surgery | within about 2 years, at −23.4% | partial regain to −16.1% at 10 years |
The drug trial is the cleanest illustration of what a plateau actually is. In a two-year study, reductions in weight "appeared to plateau around week 60 with semaglutide," and the group was at −15.6% at week 52 and −15.2% at week 1043. That is a plateau lasting a full additional year with no one doing anything wrong: the intervention was still running, the dose was unchanged, and the curve was simply finished. The surgical figures come from the Swedish Obese Subjects study, where the surgery group was down 23.4% at 2 years and 16.1% at 104.
What orders that table is not metabolic rate. It is how hard appetite pushes back — the arm of the body's defense that runs about three times larger than the expenditure side, costed in metabolic adaptation and plateaus. The intervention that blunts hunger hardest is the one whose curve stays down longest. It is also worth knowing that the underlying mathematics predicts a slow approach to any new steady weight regardless of method: the bodyweight response to a change in intake has a half-time of roughly a year5, which means flattening is the expected end of the curve rather than a fault in it.
Waiting it out has been tested, and it took eight years to read#
The largest behavioral trial ever run gives the clearest picture of what "wait and see" looks like at full length. Look AHEAD randomized 5,145 adults with overweight or obesity and type 2 diabetes to intensive lifestyle intervention or to a diabetes support and education control. The intervention group's maximum loss came at year 1 at about 8.5%, fell to 4.4% by year 4, and sat at 4.7% ± 0.2% at year 8; the control group was down 2.1% ± 0.2% at year 82. The trial's own description of the shape is that regain following maximum weight loss at year 1 itself plateaued between years 4 and 6.
Notice what did not happen. Nobody stopped supporting these participants — the intervention continued, with contact and counselling — and the line still turned and then settled. Whatever the year-1 flattening was, it lasted the remaining seven years of the study. If you are waiting for a stall to resolve without changing anything, that is the timescale the best-funded attempt to answer the question actually produced.
Two limits belong on it. Everyone in Look AHEAD had type 2 diabetes and many were taking medications that affect weight, so this is not a clean model of a healthy dieter four months into a cut. And half the intervention group still held a loss of 5% or more at year 8, which is a genuinely good outcome — the point is not that the effort failed, but that the flat part of the curve was the destination rather than a delay.
Three flat spells, three durations#
| What you are actually in | How long it lasts | What ends it |
|---|---|---|
| Fluid, glycogen and gut contents | 3–14 days | your kidneys; there is nothing to do |
| A trend too short to resolve | up to about 3 weeks | more data, not more discipline |
| A deficit that has closed | no natural end | you reopen it |
| Arrival at a weight your body defends | indefinite | a different question entirely — see plateau or set point |
The practical consequence is that the thing worth timing is not the stall. It is your response. Give a flat trend three to four weeks of seven-day averages before you touch anything, because acting inside that window mostly means acting on noise. Then act, because the base rate above says the alternative to acting is drift — and the levers, in the order their evidence supports, are laid out in how to break a plateau. If the flat stretch has arrived alongside a diet you are visibly tired of running, the thing that stalled may be the adherence rather than the physiology, which is its own diagnosis.
And if the stall has held for months while everything else in your life has stayed the same, the question worth asking is no longer how long it will last. It is whether it is still a stall — because a weight you hold effortlessly for a year is not a plateau, it is maintenance that you did not consciously choose.
FAQ#
Will a weight-loss plateau break on its own if I wait?#
Usually not, once it is real. In 667 people's weigh-in records, the trajectory that plateaued and then rose was three times as common as the one that plateaued and then fell further — 46.9% versus 15.8% of all curves1. A flat spell under three weeks is a different matter and typically clears by itself, because it is fluid rather than fat. Past that, waiting is a choice with a measured base rate, and the base rate is drift.
Is it normal to stop losing after three to six months of dieting?#
Yes, and it is close to the expected timing. Behavioral programs reliably show maximum weight loss around the end of the first year — Look AHEAD's intervention group peaked at about 8.5% at year 12 — and most of that loss has happened by the six-month mark. Arriving at a flat stretch on that schedule tells you nothing is wrong with you. It tells you the deficit you opened has closed, which is what deficits do.
Why do people on weight-loss drugs plateau later than dieters?#
Because the drug blunts the arm of the response that closes the deficit. Semaglutide's weight curve flattens around week 60 rather than month six, then holds — −15.6% at week 52 and −15.2% at week 1043. The difference between that timetable and a diet's is not metabolic rate; it is appetite. A dieter's intake creeps back up over months, while a treated patient's does so far more slowly, so the two curves flatten at different points for the same underlying reason.
Sources#
- Zhou B, Roberts SB, Das SK, Naumova EN. Weight Loss Trajectories and Short-Term Prediction in an Online Weight Management Program. Nutrients. 2024;16(8):1224.
- Look AHEAD Research Group. Eight-year weight losses with an intensive lifestyle intervention: the Look AHEAD study. Obesity (Silver Spring). 2014;22(1):5-13.
- Garvey WT, Batterham RL, Bhatta M, et al. Two-year effects of semaglutide in adults with overweight or obesity: the STEP 5 trial. Nat Med. 2022;28(10):2083-2091.
- Sjöström L, Lindroos AK, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med. 2004;351(26):2683-93.
- Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, Swinburn BA. Quantification of the effect of energy imbalance on bodyweight. Lancet. 2011;378(9793):826-37.
