A tirzepatide plateau is the point where the scale stops moving even though nothing about the treatment has changed. It is a normal feature of the trial curves, not a sign of failure.
The useful question is when it usually happens, and what the trials show about pushing past it. How fast the loss starts is covered in how long tirzepatide takes to work.
What the trial curves do
SURMOUNT-1 randomized 2,539 adults with obesity, or overweight with a weight-related condition, to 5, 10 or 15 mg weekly or placebo for 72 weeks [1]. The first 20 weeks were dose escalation.
| Dose | Week 72 (full trial) | Week 176 (prediabetes subgroup) |
|---|---|---|
| 5 mg | −15.0% (95% CI −15.9 to −14.2) | −12.3% |
| 10 mg | −19.5% (95% CI −20.4 to −18.5) | −18.7% |
| 15 mg | −20.9% (95% CI −21.8 to −19.9) | −19.7% |
| Placebo | −3.1% (95% CI −4.3 to −1.9) | −1.3% |
The two columns are not the same population, so they are not a clean before-and-after. They still show the shape.
Between week 72 and week 176, more than two further years of treatment, the mean figures did not improve. That is the plateau, at the level of a whole trial.
What the plateau preserved is the loss itself. Nobody in those groups drifted back toward baseline while treatment continued.
A stall is not always a plateau
SURMOUNT-4 gave 783 adults open-label tirzepatide for 36 weeks, reaching a mean 20.9% reduction [3]. The arm that continued for another 52 weeks lost a further 5.5%.
So at 36 weeks, near the point many people describe stalling, the average participant still had 5% to come. A slow month is not proof the curve has finished.
Dose is part of that. The label starts everyone at 2.5 mg for four weeks, then 5 mg [4]. After that it allows increases of 2.5 mg every four weeks, up to a maximum of 15 mg weekly.
Someone who settled at 5 mg is on a dose whose trial average was 15.0%, not 20.9%. The ladder is set out in tirzepatide dosage.
Or it may be the response you were going to get
Not everyone responds the same way, and the difference shows up early.
A 2026 review of response heterogeneity defines non-response as losing under 5% of body weight [6]. It puts that at around 10% of participants in controlled trials of obesity without diabetes. Real-world cohorts suggest a higher rate.
Its conclusion on prediction is blunt. No single baseline characteristic reliably predicts who responds, and early on-treatment response is the most defensible signal available.
Genetic markers are not ready. The review describes a large study of variants in the drug-target genes as hypothesis-generating rather than a basis for genotyping anyone.
What to raise with a prescriber
Ask how long you have been at your current dose, and whether the ladder has further to go.
Ask what your loss was at week 12 and week 20. Early response is the best predictor anyone currently has.
Ask whether the goal has changed. Holding a 20% loss is an outcome in itself, and the trials show it takes continued treatment.
And ask what the loss is made of, since lean mass is part of it. That is covered in GLP-1 muscle loss.
Putting it together
Weight loss on tirzepatide slows because the curve is meant to slow. In the trials the mean reached its level somewhere after the first year and stayed there while treatment continued.
Before concluding a plateau has arrived, the things worth checking are dose, time at that dose and what the first 20 weeks looked like.
For the full range of results by dose, see how much weight you can lose on tirzepatide. To compare prescribers by monthly cost, see the compounded tirzepatide rankings, and for the rest of this goal, the weight hub.