SURMOUNT Trial Results Explained
The SURMOUNT trial family produced Zepbound's approval. Here's what SURMOUNT-1, 2, 3, and 4 actually showed — and what the data does and doesn't generalize to.
Updated May 6, 2026 · 6 min read
The SURMOUNT program is the trial family Eli Lilly ran to get Zepbound approved for chronic weight management. Four trials, four populations, four overlapping questions about how tirzepatide performs in different contexts. The results are some of the most striking weight-loss data ever published — and worth reading carefully, because the headlines (22.5% weight loss!) compress a lot of nuance.
SURMOUNT-1: the headline number
Population: 2,539 adults with obesity (BMI ≥ 30) or overweight + at least one weight-related comorbidity, without type 2 diabetes. Duration: 72 weeks (about 17 months). Design: Randomized 1:1:1:1 to placebo, 5 mg, 10 mg, or 15 mg tirzepatide weekly. All arms received lifestyle counseling.
| Arm | Mean weight loss at week 72 | % achieving ≥ 5% loss | % achieving ≥ 20% loss |
|---|---|---|---|
| Placebo | -2.4% | 35% | 1.5% |
| Tirzepatide 5 mg | -16.0% | 85% | 30% |
| Tirzepatide 10 mg | -21.4% | 89% | 50% |
| Tirzepatide 15 mg | -22.5% | 91% | 57% |
This is the trial that produced the "22.5%" number you'll see everywhere. A few things worth noting:
- The placebo arm lost 2.4% — that's the lifestyle counseling effect alone. Tirzepatide's incremental effect over placebo is roughly 14–20 percentage points, depending on dose.
- Even 5 mg outperforms semaglutide's top dose in head-to-head comparison (semaglutide 2.4 mg in STEP-1: ~14.9% over 68 weeks).
- The 10 → 15 mg jump adds about 1 percentage point of average weight loss. Real, but small. See when to step up for the implications.
SURMOUNT-2: T2D + obesity
Population: 938 adults with both type 2 diabetes and obesity. Duration: 72 weeks. Design: Placebo vs 10 mg vs 15 mg tirzepatide.
| Arm | Mean weight loss at week 72 |
|---|---|
| Placebo | -3.2% |
| Tirzepatide 10 mg | -13.4% |
| Tirzepatide 15 mg | -15.7% |
Notice the gap. SURMOUNT-1 (no diabetes): 22.5% on 15 mg. SURMOUNT-2 (with diabetes): 15.7% on 15 mg. Patients with T2D lose less weight on the same dose. This holds across the GLP-1 class and isn't unique to tirzepatide.
The mechanism is partly metabolic — insulin resistance dampens GLP-1/GIP response — and partly behavioral. T2D patients often have different baseline eating patterns and may be on other medications affecting weight. See T2D vs weight loss for the deeper story.
A1c reductions in SURMOUNT-2 were strong — averaging 2+ percentage points on 15 mg — confirming that the drug does both jobs even when the population is selected for one.
SURMOUNT-3: after intensive lifestyle
Population: Adults with obesity who completed a 12-week intensive lifestyle intervention before randomization. Duration: 72 weeks of randomized treatment after the lifestyle phase. Design: All participants did the lifestyle program first; then placebo vs tirzepatide.
The setup matters: this trial answered "does tirzepatide add weight loss on top of what serious lifestyle work already produced?"
| Phase | Lifestyle alone (12 weeks) | Subsequent randomization (72 weeks) |
|---|---|---|
| Mean weight loss | -6.9% | Placebo: +0.7% (regain); Tirzepatide: -21.1% additional |
Two takeaways:
- Lifestyle alone produces about 7% loss in 12 weeks — meaningful, and consistent with intensive-program literature.
- Adding tirzepatide produces an additional 21% loss on top of that. The placebo arm slowly regained, while the tirzepatide arm kept descending.
This is the strongest evidence that tirzepatide isn't just substituting for behavior change — it's adding effect on top of it.
SURMOUNT-4: what happens when you stop
Population: Adults who had completed 36 weeks of open-label tirzepatide titration (most reaching 10 or 15 mg). Duration: 52 weeks of randomized continuation vs withdrawal. Design: Continue tirzepatide vs switch to placebo.
| Arm | Weight change during 52-week randomized phase |
|---|---|
| Continue tirzepatide | -5.5% additional loss |
| Switch to placebo | +14% regain |
The withdrawal arm regained the majority of what they had lost. Stopping tirzepatide is roughly equivalent to stopping the metabolic effect — appetite returns, eating returns to pre-treatment patterns, and weight follows.
This is the trial that crystallized the "obesity is a chronic disease" framing for GLP-1 therapy. Tirzepatide isn't a 12-month course; it's a long-term treatment, with the same underlying logic as antihypertensives.
What the placebo arms tell us
A consistent and underappreciated finding: placebo arms in SURMOUNT lost 2–3% of body weight despite no active drug. This is the lifestyle counseling effect, plus regression to the mean, plus the Hawthorne effect of being in a structured trial.
It matters because:
- Real-world tirzepatide users don't get the trial-level lifestyle counseling. Some of the 14-percentage-point gap between drug and placebo in SURMOUNT-1 reflects the structured trial environment.
- A real-world user with no behavior change might lose less than the 16% trial number for 5 mg — because the placebo "floor" they're starting from is closer to 0% than to -2.4%.
- Conversely, a real-world user with active behavior change might exceed the trial numbers. SURMOUNT-3 essentially showed this.
What the data does and doesn't generalize to
| Generalizes well to | Generalizes weakly to |
|---|---|
| Adults aged 18–75 with BMI ≥ 27 | Adolescents (separate trials, smaller data) |
| Patients on no other weight-loss drugs | Patients on combination therapy |
| Patients without serious cardiovascular disease | Patients with recent MI or unstable cardiovascular conditions |
| 72-week timeframe | Multi-year outcomes (longer-term data still maturing) |
| Brand-name tirzepatide on standard titration | Compounded or off-protocol use |
The adolescent question is partially answered by SURMOUNT-ADOLESCENTS, an ongoing program. Pediatric labels haven't been added yet.
The cardiovascular outcomes question is the subject of SURPASS-CVOT (a separate ongoing trial in T2D). Until those data publish, the cardiovascular protective story is plausible based on weight and A1c effects but not formally established the way semaglutide's is via SELECT.
SURMOUNT vs STEP (semaglutide)
For comparability, the semaglutide STEP-1 trial:
| Trial / dose | Weight loss at study end |
|---|---|
| STEP-1 (semaglutide 2.4 mg, 68 weeks) | -14.9% |
| SURMOUNT-1 (tirzepatide 15 mg, 72 weeks) | -22.5% |
| SURMOUNT-1 (tirzepatide 5 mg, 72 weeks) | -16.0% |
Tirzepatide at its lowest therapeutic dose matches semaglutide at its highest approved dose. That's the comparison that drives most of the "tirzepatide is more effective" framing — and was confirmed in the SURMOUNT-5 head-to-head trial. See tirzepatide vs semaglutide.
What the trials don't answer
Honest gaps in the SURMOUNT data:
- What happens past 2 years? The trials largely ended at 72–88 weeks. Long-term users are answering this question in real time.
- Who responds best? Subgroup analyses suggest younger, female, and lower-baseline-BMI patients lose proportionally more, but the effect is present across most groups.
- What's the optimal maintenance strategy? SURMOUNT-4 showed withdrawal causes regain, but didn't test step-down protocols or intermittent dosing.
- How does it interact with menopause, pregnancy planning, eating disorders, or specific psychiatric conditions? Limited trial data; clinician judgment fills the gap.