Semaglutide Week by Week: When the Weight Actually Comes Off

If you are three weeks in and the scale has barely moved, nothing has gone wrong. The starting dose of semaglutide is not a treatment dose — it exists so your gut can meet the drug slowly — and in the trial designed to measure the climb, participants spent 20 weeks getting to the full 2.4 mg and had lost about 10.6% of their body weight by the time they arrived.^[3] The steep part of the curve comes after that. Here is what the four STEP trials actually recorded, week by week, and what happened to the people who stopped.

The first weeks are a dose ladder, not a result

Semaglutide is started low and raised in steps, and the reason is tolerance rather than caution about efficacy. Beginning anyone at 2.4 mg would produce nausea severe enough that most would quit before the drug had a chance to work. So the label builds in a climb, and the early weeks belong to that climb.

This matters for expectations more than anything else on this page. A reader who has been told “people lose 15% on this drug” and is standing on a scale in week 4 is comparing themselves against a 68-week endpoint while still on an introductory dose. Those are not the same measurement.

What 68 weeks produced

STEP 1 is the trial every later result is measured against. It enrolled 1,961 adults living with obesity and without diabetes, gave them weekly semaglutide 2.4 mg or placebo, and ran for 68 weeks. The treated group finished at −14.9% of body weight against −2.4% on placebo.^[1]

The average conceals a wide spread, and the spread is the more useful fact. Just over half the treated group lost at least 15% of their body weight. On placebo, about one in twenty did.^[1] Some people in that trial did far better than 14.9% and some did considerably worse, which is why no honest version of this page can tell you what you personally will lose.

The curve is steep in the middle and flattens at the end

Putting STEP 4 and STEP 1 side by side gives the shape. Twenty weeks of dose escalation produced about 10.6%. Continuing at the full dose for another 48 weeks added roughly 7.9% more.^[3] So the middle stretch — full dose, months five through twelve — is where the drug does its heaviest work, and the last months of a 68-week course add less than the months before them.

A plateau late in the first year is therefore the expected shape of the curve rather than a sign the drug has stopped working. That is worth saying plainly, because a plateau is one of the most common reasons people stop taking a GLP-1, and the trials suggest it is the normal end of the ramp rather than a failure.

The drug and the program add up

STEP 3 asked whether semaglutide and serious behavioral therapy overlap or stack. Everyone in it got the intensive program — roughly 30 visits and a restricted calorie target — and half also got the drug. The drug arm finished at 16.0%; the therapy-only arm at 5.7%.^[2] They add.

That finding does real work outside the clinic. Prior-authorization criteria routinely require a documented lifestyle program alongside the prescription, and STEP 3 is the evidence that the requirement is not merely bureaucratic. If you are assembling that documentation, our guide to getting a GLP-1 covered sets out what payers ask to see.

What happens when it stops

This is the part of the timeline least often described to people starting out, and it changes what the whole course means.

STEP 4 ran the experiment inside the trial. After the 20-week run-in, half the participants were moved to placebo while the other half continued. The continuers lost a further 7.9%. The switchers regained 6.9% — giving back most of what the run-in had achieved, within the same trial, while everything else about their care stayed the same.^[3]

The STEP 1 extension followed 327 people for a year after treatment ended altogether. Having lost 17.3%, they regained 11.6 percentage points of it, finishing about 5.6% below where they started. Blood pressure, lipids, A1C and inflammatory markers drifted back toward baseline alongside the weight.^[4]

Two-thirds of the loss came back within a year of stopping, on a drug that had been working.

That result is why obesity medicine now describes these drugs as treatment for a chronic condition rather than a course with an end date. It is also the fact most worth knowing before the first injection, because it turns the question from “how much will I lose” into “what happens when I can no longer get this” — which for a great many people is a question about cost and supply rather than biology. We keep a dated price for every seller on the price tracker, and what happens to the body afterward is covered in what happens when you stop a GLP-1.

What this timeline cannot tell you

• It is a group average, not a forecast. Every figure above is a mean across hundreds or thousands of people, and individuals scattered widely around it in both directions.
• Trial conditions are not ordinary conditions. Participants had structured support, free medication and regular contact with a study team. Adherence in a trial is not adherence in a life.
• It describes brand semaglutide at the labeled dose. A compounded vial is a different product with its own questions, taken up in compounded semaglutide versus brand.
• It is not a tirzepatide timeline. Zepbound and Mounjaro climb a different ladder and were measured in different trials. Their curve is a separate piece of work, not a paragraph borrowed from this one.

One thing we could not establish from the published record: how much of the early-weeks loss is attributable to the drug rather than to the dietary changes people make in the first month of any serious attempt. STEP 3 shows the two add over 68 weeks, but it does not resolve the first four.