For four years the honest answer to "which one works better" was that nobody had measured it. Semaglutide had its trials, tirzepatide had its trials, and the two sets of numbers were not comparable to each other. That changed in May 2025, when SURMOUNT-5 was published: 751 adults with obesity and without diabetes, randomly assigned to one drug or the other, weighed at 72 weeks.
This page reports what that trial found and, in the same detail, what about its design limits the finding. It does not say which medicine anyone should take — that belongs with a prescriber, for the reasons our editorial standards set out. Our cost page carries the current prices for both drugs, and our retatrutide vs tirzepatide page covers the molecule that is being lined up behind them. For the tirzepatide side on its own — what it achieved over 88 weeks, and the odd fact that the commonest thing reported about it to the FDA is not a side effect at all — there is a longer look at the drug's own record.
The head-to-head result
SURMOUNT-5 was a phase 3b trial run by Eli Lilly (NCT05822830). Adults with obesity but without type 2 diabetes were randomised 1:1 to the maximum tolerated dose of tirzepatide (10 mg or 15 mg) or the maximum tolerated dose of semaglutide (1.7 mg or 2.4 mg), injected once weekly for 72 weeks. The primary endpoint was percent change in weight.
| Endpoint at week 72 | Tirzepatide | Semaglutide |
|---|---|---|
| Mean weight change | −20.2% (95% CI −21.4 to −19.1) | −13.7% (95% CI −14.9 to −12.6) |
| Mean waist-circumference change | −18.4 cm (95% CI −19.6 to −17.2) | −13.0 cm (95% CI −14.3 to −11.7) |
| Reached −10%, −15%, −20%, −25% | More likely, at each threshold | Reference |
| Most common adverse events | Gastrointestinal, mostly mild to moderate, mostly during escalation | Gastrointestinal, mostly mild to moderate, mostly during escalation |
Both differences were significant at P<0.001 (Aronne et al., N Engl J Med 2025;393:26-36). That is the result, and it is the result almost every article about these two drugs now cites. It is also the only head-to-head trial the class has: every other product's figure comes from its own separate study, measured against its own placebo arm. Its side-effect counts, set beside the four other head-to-head trials in the class, are compared on our fewest-side-effects page.
Four things about the trial that decide what the result is worth
Nothing below claims the trial was wrong. Each item is a design fact stated in the paper itself, and each one changes how far a 6.5-percentage-point gap should be carried.
It was open-label. Every participant and investigator knew which drug was in the pen. The two medicines use different devices on different escalation schedules, so blinding would have meant dummy injections in both arms; the trial did not do that. Weight-loss outcomes depend heavily on what participants do between visits, and expectation is not a neutral input.
It was funded by one side. Lilly makes tirzepatide, designed and paid for the trial, and eight of the twelve named authors are Lilly employees. Industry funding does not make a result false — SURMOUNT-5 is the only randomised comparison that exists, and it would not exist otherwise — but a reader is entitled to know that the drug that won was the sponsor's.
The comparator arm could finish below the approved dose. Both arms were titrated to the maximum dose the participant tolerated. For tirzepatide that meant 10 mg or 15 mg, both approved maintenance doses. For semaglutide it meant 1.7 mg or 2.4 mg — and 1.7 mg is an escalation step, not the 2.4 mg dose approved for weight management. The design is defensible as a picture of real prescribing, where not everyone reaches the top dose. It is not the same trial as one comparing 15 mg with 2.4 mg.
It is one trial of 751 people. SURMOUNT-1 enrolled 2,539 and STEP 1 enrolled 1,961. A single head-to-head of this size settles the direction of the difference much better than it settles its size. Peer correspondence followed: three letters, from Leiden University Medical Center and from GFO Clinics Troisdorf, were published with the authors' reply in N Engl J Med 2025;393(16):1654-1656.
What each drug did against placebo
These are the trials that established each medicine. They are listed here for context, and they are not comparable to each other — different years, populations, durations and estimands.
| Trial | Drug | Participants | Duration | Mean weight change | Placebo |
|---|---|---|---|---|---|
| STEP 1 | Semaglutide 2.4 mg weekly | 1,961 | 68 weeks | −14.9% | −2.4% |
| SURMOUNT-1 | Tirzepatide 5 / 10 / 15 mg weekly | 2,539 | 72 weeks | −15.0% / −19.5% / −20.9% | −3.1% |
In STEP 1, 86.4% of the semaglutide group lost at least 5% of body weight, 69.1% lost at least 10% and 50.5% lost at least 15%, against 31.5%, 12.0% and 4.9% on placebo (Wilding et al., N Engl J Med 2021;384:989-1002). Discontinuation for gastrointestinal events was 4.5% on semaglutide and 0.8% on placebo.
What separates them apart from the numbers
Receptors. Semaglutide is a GLP-1 receptor agonist. Tirzepatide activates the GIP receptor as well — which is why only one of the two is a GLP-1 in the strict sense. Whether that second receptor is the reason for the difference in SURMOUNT-5 is not established by SURMOUNT-5, which measured weight, not mechanism.
Neither is a novel molecule. Semaglutide's own label puts it at 94% sequence homology with the GLP-1 the gut already makes, and the engineering that separates them is about survival rather than strength — the body's own version has a half-life of 1.5 to 2 minutes against semaglutide's week.
The molecule semaglutide displaced. Before this pairing became the comparison people searched for, semaglutide had already been tested against the older daily injection it replaced, and that trial is the most lopsided head-to-head in the category: −16.4% against −6.4% at week 68. It is useful context for the margins below, which are far narrower.
The brands are not the same comparison. Sold as Zepbound and Wegovy, the two products carry different approved indications, reach different age groups and price along different curves — none of which follows from the molecules or from SURMOUNT-5. That is a separate question and it has its own page.
The same two molecules are also sold as Mounjaro and Ozempic for type 2 diabetes, where the only direct trial is SURPASS-2 rather than SURMOUNT-5 and the heart claims on the two labels rest on quite different evidence. That pair has its own comparison.
Formulations. Both are once-weekly injections for weight management, sold as Wegovy and Zepbound. Semaglutide now also exists as a daily tablet: FDA approved Wegovy tablets under NDA 218316 on 2025-12-22, and the trial behind it, OASIS 4, reported −13.6% at 64 weeks in 205 people on 25 mg. Tirzepatide has no oral form — the tablet competing with it is a different molecule again, orforglipron, which we compare with tirzepatide on its own page. Our GLP-1 pills page covers both oral options and their prices.
Price. Read from the manufacturers on 2026-09-06:
| Self-pay price per month | |
|---|---|
| Zepbound 2.5 mg / 5 mg / 7.5–15 mg | $299 / $399 / $449 |
| Wegovy pen, most doses | $349 ($199 for a new patient's first two fills) |
| Wegovy pill 1.5 mg / 4 mg / 9 mg and 25 mg | $149 / $199 / $299 |
Both manufacturers advertise commercial-insurance savings cards at "as little as $25" a month with caps. The cost hub keeps the full table, including Medicare terms, with the date each figure was read.
What we would want next
A blinded trial comparing 15 mg of tirzepatide with 2.4 mg of semaglutide, run by neither manufacturer, reporting outcomes past 72 weeks. Nobody has announced one. Until then the honest summary is narrow: in one open-label, industry-funded, maximum-tolerated-dose trial of 751 people, tirzepatide produced more weight loss than semaglutide at 72 weeks, and the direction of that finding is better supported than its exact size.
Sources
- Aronne LJ et al. Tirzepatide as Compared with Semaglutide for the Treatment of Obesity. N Engl J Med 2025;393(1):26-36. PubMed 40353578 · NCT05822830
- Correspondence on SURMOUNT-5, N Engl J Med 2025;393(16):1654-1656. PubMed 41124642, 41124643, 41124644
- Wilding JPH et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. N Engl J Med 2021;384(11):989-1002. PubMed 33567185
- Wharton S et al. Oral Semaglutide at a Dose of 25 mg in Adults with Overweight or Obesity. N Engl J Med 2025;393(11):1077-1087. PubMed 40934115
- Manufacturer self-pay prices, read 2026-09-06: NovoCare Wegovy savings and Lilly's Zepbound coverage page, as recorded on our cost hub.
- FDA approval records via the openFDA Drugs@FDA API, read 2026-09-06 (Wegovy tablets, NDA 218316, original approval 2025-12-22).