How Fast Does Semaglutide Work? A Week-by-Week Reality Check

“How fast does semaglutide work” is one of the most-typed questions about this drug, and most of the answers online are charts of pounds-per-week that nobody measured. The clinical trials did not report weight that way. They reported it at week 68 or week 104, with a curve in between and a handful of well-documented waypoints.

This page walks through the waypoints that actually exist in the FDA-approved labeling and the published trials, explains what happens between them, and is honest about the parts nobody has measured.

Two ground rules first. Nothing here is a dosing schedule or a target — dose and timing decisions belong to a prescriber, and the approved labeling is the authority on how these products are used. And every number is a group average from a trial. The spread around those averages is enormous, and a later section covers it.

About the label material on this page. Where this article describes dosing schedules, escalation, missed doses, storage or warnings, it is summarizing the FDA-approved prescribing information and Instructions for Use for Wegovy and Zepbound as published on DailyMed. Those documents are the authority, they are revised periodically, and they differ by product and by presentation (pen, prefilled syringe, single-dose vial, multi-dose vial). Follow the Instructions for Use that came with your own prescription, and take any question about your own dose, timing or supply to your prescriber or pharmacist rather than to this page.

What actually happens in the first four weeks?

In the STEP trials, semaglutide was started low and stepped up about every four weeks over a 16-week escalation before reaching the 2.4 mg maintenance dose [1][2]. The 7.2 mg dose (Wegovy HD, FDA-approved March 2026) is reached only after at least four further weeks at 2.4 mg, so about 20 weeks at the earliest [1]. The tablet (Wegovy tablets, FDA-approved December 2025) escalates over 90 days, moving 1.5 mg to 4 mg to 9 mg to 25 mg in 30-day blocks [1].

That escalation exists to reduce nausea and vomiting, not to produce weight loss. This is the single biggest mismatch between what people expect and what the trials did: most people start counting results in week one, while trial participants spent that whole first month on a small fraction of the dose that generated the published numbers.

What people commonly describe in this window is a change in appetite — feeling full sooner, thinking about food less — often before the scale reflects much. That matches the mechanism. In a 20-week randomized trial of 72 adults that measured what people actually ate at a self-served lunch, the semaglutide 2.4 mg group consumed 35% fewer calories than placebo and reported less hunger, more fullness, and fewer and weaker food cravings [3].

Side effects also cluster here and at each step up. The label lists nausea, vomiting, constipation, abdominal pain, fatigue, headache and dizziness among the most common reactions, and in the 7.2 mg trials nausea was reported by 39% of that group against 13% on placebo [1].

Weeks 5 to 16: what is happening while the dose climbs?

Weight is coming off during this stretch, but the curve is still shallow, because the full dose has not been reached for most of it. Each escalation step often brings a fresh round of gastrointestinal symptoms that then settles.

This window is also where the biggest gap between trial data and real life opens up. In STEP 1, only 4.5% of participants stopped because of gastrointestinal events [2]. In the real world, a US health-system cohort of 7,881 adults found that people who discontinued within three months had lost an average of just 3.6% of body weight at one year, versus 11.9% among those who did not discontinue [4].

A note on what else was happening in the trials during these weeks: participants were not only taking a drug. In Wegovy’s pivotal trials, everyone received instruction on a reduced-calorie diet of roughly a 500 kcal/day deficit and counseling to reach at least 150 minutes of physical activity per week, starting with the first dose and continuing throughout [1]. A 2026 review of 42 phase 3 obesity-medication trials found that package — roughly a 500 kcal deficit, at least 150 minutes a week, food and activity diaries, and counseling at least every three months and often monthly — was the norm across the whole field [5]. The published results are results from that combination.

Week 20: the clearest milestone in the whole literature

STEP 4 is the trial that gives the best early data point, because it was designed around one. All 902 enrolled adults took semaglutide for a 20-week run-in that included the 16-week escalation. Only then were they randomized. At the end of that run-in, mean weight loss was about 10.6% of body weight [6].

For someone starting at 235 pounds, that is roughly 25 pounds in five months. It is far more than a typical diet produces in that time. It is also well short of the 14.9% headline from STEP 1, which took another 48 weeks to reach [2].

The same trial reports something else worth knowing: 11% of the original 902 never made it to randomization, most commonly because of adverse reactions (5.3%) [1]. Reaching the maintenance dose is not automatic.

Does a slow start mean it will not work?

This is the most useful question in the whole timeline, and STEP 4 answered it directly in a post hoc analysis.

Researchers split participants by whether they had lost at least 5% by week 20. Among the 88.0% who had, mean change from week 0 to week 68 was -19.7%. Among those who had not, continuing semaglutide still produced -6.4% on average, versus -0.3% for those switched to placebo [7].

Statistically: being an early responder predicted reaching 5% at week 68 with a positive predictive value of 96.4%. Being a non-responder at week 20 had a negative predictive value of only 42.9% [7].

In plain English — a strong early result almost always means a strong final result, but a slow start tells you very little. The same pattern shows up with tirzepatide: in SURMOUNT-1, 18% of adherent participants lost less than 5% by week 12, and 90% of that group had reached 5% by week 72, taking on average about 25 weeks to get there [8].

What do the 68- and 72-week numbers actually say?

Here is what the FDA-approved Wegovy labeling reports, using the intention-to-treat population — the more conservative accounting, which is why these run slightly lower than the figures in press releases [1]:

Trial population	Mean weight change	Placebo
Obesity or overweight, no diabetes (68 wk, 2.4 mg)	-14.9%	-2.4%
Type 2 diabetes (68 wk, 2.4 mg)	-9.6%	-3.4%
With intensive lifestyle therapy (68 wk, 2.4 mg)	-16.0%	-5.7%
Obesity, no diabetes (72 wk, 7.2 mg)	-18.8%	-3.9%
Obesity, no diabetes (72 wk, 2.4 mg comparator)	-15.5%	-3.9%
Obesity or overweight (64 wk, 25 mg oral tablet)	-13.6%	-2.4%

Two things stand out. First, results in people with type 2 diabetes are consistently smaller — this is a well-replicated finding, not a fluke [9]. Second, the highest dose buys a few more percentage points and takes longer to reach.

When does it plateau?

In STEP 1 the curve flattened toward the end of the 68 weeks [2]. STEP 5 extended treatment to 104 weeks and found a mean of 15.2%, against STEP 1’s 14.9% at 68 weeks — essentially flat across the entire second year [10].

So the honest shape of the semaglutide curve is: shallow for the first few months while the dose climbs, steepest through the middle of the first year, and flattening somewhere around 12 to 15 months.

That flattening is not the drug wearing off. As body mass falls, so does the energy the body burns, and a new balance point is reached. STEP 4 showed what the drug is still doing at that point: people who continued lost a further 7.9% between weeks 20 and 68, while those switched to placebo gained 6.9% [6]. The flat part of the curve is the drug holding a line, not doing nothing.

Why are real-world results smaller?

Because the trials measured people who reached and stayed on the full dose, with structured counseling, for 68 weeks.

The 7,881-patient cohort mentioned earlier found a mean one-year reduction of 8.7% across everyone who started, with 80.8% on low maintenance doses [4]. A separate comparison across two health systems put real-world semaglutide total weight loss at 5.4% at one year on an intention-to-treat basis, rising to 7.2% among people with a full year of continuous orders [11].

Persistence is the main variable. Reviews of real-world data find continuation often falling below 60% by 12 to 24 months, driven mostly by gastrointestinal intolerance and cost [12].

None of that makes the trial numbers wrong. It means the trial numbers describe what happens when treatment continues at an effective dose — which is exactly the thing readers should be planning for with their prescriber, rather than treating as automatic.

What does the spread around the average look like?

The mean hides a lot. In STEP 1, at week 68 [2]:

• 86.4% lost at least 5% (placebo: 31.5%)
• 69.1% lost at least 10% (12.0%)
• 50.5% lost at least 15% (4.9%)
• 32.0% lost at least 20% (1.7%)

Which also means roughly one in seven people on semaglutide did not reach 5%. Across the STEP program as a whole, 32-39.6% were “super responders” losing more than 20%, and 10.2-16.7% lost less than 5% [9].

Two predictors of a smaller result have been identified consistently: having type 2 diabetes, and being male [9]. The rest of the variation is not yet explained, and precision-medicine work aimed at predicting response is ongoing.

What about pounds per week?

There is no published pounds-per-week figure for semaglutide, and any chart offering one is extrapolating. The nearest defensible arithmetic is to divide a milestone by its timeframe: about 10.6% over 20 weeks in STEP 4 works out to roughly 0.5% of body weight per week during the fastest phase, slowing considerably after that [6].

That is a group average during the steepest part of a curve, applied backwards. It is not a target, and using it to judge any individual week would be a mistake — home-scale readings shift with fluid, sodium, glycogen, bowel contents and the menstrual cycle, which is why the trials analyzed 12-week windows rather than weekly readings.

What happens if you stop?

The STEP 1 extension followed 327 participants for a year after both the drug and the structured lifestyle program ended. They had lost 17.3% at week 68 and regained 11.6 percentage points by week 120, ending 5.6% below their starting weight. Nearly half (48.2%) still had at least 5% loss at that point, down from 86.4% at the end of treatment, and most cardiometabolic improvements had reverted [13].

A 2026 systematic review in eClinicalMedicine screened 48 studies and modeled the shape from six randomized trials (3,236 participants): about 60% of the lost weight comes back within a year, and the model projects regain leveling off at around 75% of what was lost [14].

The short version

• The first month is a titration step, not a results window.
• Five months in, the best-documented average is about 10.6%.
• The curve is steepest in the middle of the first year and flat by 12 to 15 months.
• A slow start predicts very little; an early strong response predicts a lot.
• The label averages (14.9% at 2.4 mg, 18.8% at 7.2 mg) describe people who reached and stayed on the dose.
• Real-world one-year averages are closer to 8.7%, driven mostly by stopping early and low doses.

If your own trajectory does not match any of this, that is a conversation to have with a healthcare provider — someone who can look at dose, timing, other medications and everything else going on — rather than a reason to change anything on your own.

Sources

1. Wegovy (semaglutide) prescribing information, Novo Nordisk, via DailyMed, June 2026. https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=ee06186f-2aa3-4990-a760-757579d8f77b
2. Wilding JPH et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. NEJM 2021. https://www.nejm.org/doi/full/10.1056/NEJMoa2032183
3. Friedrichsen M et al. The effect of semaglutide 2.4 mg once weekly on energy intake, appetite, control of eating, and gastric emptying in adults with obesity. Diabetes, Obesity and Metabolism 2021;23(3):754-762. https://doi.org/10.1111/dom.14280
4. Gasoyan H et al. Changes in weight and glycemic control following obesity treatment with semaglutide or tirzepatide by discontinuation status. Obesity 2025. https://doi.org/10.1002/oby.24331
5. A Review of Lifestyle Interventions Provided as an Adjunct to Obesity Management Medications. Obesity Science & Practice 2026. https://doi.org/10.1002/osp4.70162
6. Rubino D et al. Effect of Continued Weekly Subcutaneous Semaglutide vs Placebo on Weight Loss Maintenance (STEP 4). JAMA 2021. https://jamanetwork.com/journals/jama/fullarticle/2777886
7. Mosenzon O et al. Clinically-Relevant Weight Loss is Achieved Independently of Early Weight Loss Response (STEP 4 post hoc). 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC8265765/
8. Ard JD et al. Weight reduction over time in tirzepatide-treated participants by early weight loss response: SURMOUNT-1 post hoc. Diabetes Obes Metab 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12326891/
9. Tzoulis P, Baldeweg SE. Semaglutide for weight loss: unanswered questions. 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC11188346/
10. Garvey WT et al. Two-year effects of semaglutide in adults with overweight or obesity: the STEP 5 trial. Nature Medicine 2022. https://www.nature.com/articles/s41591-022-02026-4
11. Real-World Effectiveness of Semaglutide and Tirzepatide Compared With Bariatric Surgery. Obesity 2026. https://doi.org/10.1002/oby.70246
12. Persistence in GLP-1-Based Therapy Among Adults with Obesity and Type 2 Diabetes. Curr Diab Rep 2026. https://doi.org/10.1007/s11892-026-01639-0
13. Wilding JPH et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: the STEP 1 trial extension. Diabetes Obes Metab 2022. https://pmc.ncbi.nlm.nih.gov/articles/PMC9542252/
14. Trajectory of weight regain after cessation of GLP-1 receptor agonists: a systematic review and nonlinear meta-regression. eClinicalMedicine 2026. https://doi.org/10.1016/j.eclinm.2026.103796