Evidence review

Semaglutide vs tirzepatide: what the head-to-head trials found

By David Chen·Editor·Reviewed September 2026

Semaglutide’s results, side by side

Read the table across, one trial at a time. The two trials enrolled different people and used different semaglutide doses, so the rows answer different questions.

Most other comparisons of the two drugs set one trial’s result beside another’s, and the difference between the trials gets read as a difference between the drugs. These two are the exception, which is why they are the only ones on this page.

SURMOUNT-5: semaglutide at its weight-loss dose

Adults with obesity but without type 2 diabetes.

Phase 3b, open-label, 72 weeks. Participants were assigned 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), once weekly by subcutaneous injection.

This is the fairer test for semaglutide as a weight-loss drug, because its arm used the maximum tolerated dose of 1.7 or 2.4 mg, the maintenance doses on the Wegovy label. On that dose, mean weight fell 13.7% at week 72 and waist circumference by 13.0 cm. The tirzepatide arm lost 20.2% and 18.4 cm. The gap, about six and a half percentage points of body weight, was statistically significant, and more people on tirzepatide reached losses of at least 10%, 15%, 20% and 25%.

The 2.4 mg dose has its own placebo-controlled trial, STEP 1, which is the better place to see what semaglutide does on its own. This trial answers a narrower question: which of the two loses more, given the same people and the same 72 weeks.

SURPASS-2: semaglutide at a diabetes dose

Adults with type 2 diabetes. Mean baseline glycated hemoglobin was 8.28%, mean age 56.6 years and mean weight 93.7 kg.

Phase 3, open-label, 40 weeks. Participants were assigned 1:1:1:1 to tirzepatide 5 mg, 10 mg or 15 mg, or to semaglutide 1 mg, once weekly.

The semaglutide arm here was 1 mg, an Ozempic dose for type 2 diabetes and well below the 2.4 mg used for weight loss. On it, glycated hemoglobin fell 1.86 percentage points. Every tirzepatide dose did better, by 0.15 points at 5 mg up to 0.45 points at 15 mg, and tirzepatide was both noninferior and superior at all three. Weight also fell further on tirzepatide, by an estimated 1.9, 3.6 and 5.5 kg more than on semaglutide.

So this trial compares one of semaglutide’s diabetes doses with the full range of the other drug. It is a valid answer to the question it asked, and not a ceiling-to-ceiling comparison. SUSTAIN 7 looks at the 1 mg dose on its own terms.

Side effects on semaglutide in these trials

The most common adverse events in both trials, on both drugs, were gastrointestinal and mostly mild to moderate. SURPASS-2 gives the numbers. On semaglutide 1 mg, 18% reported nausea, 12% diarrhea and 8% vomiting; on tirzepatide the ranges across doses were 17% to 22%, 13% to 16% and 6% to 10%. Serious adverse events were reported in 3% on semaglutide and 5% to 7% on tirzepatide.

SURMOUNT-5’s abstract does not break out rates by drug. It says the gastrointestinal events occurred mainly during dose escalation. The Wegovy label’s own placebo-controlled rates are on semaglutide side effects.

How the trials were run

Both were open-label: participants and investigators knew which drug each person was taking. Both were funded by Eli Lilly, which makes tirzepatide, and both list company employees among their authors. Neither fact overturns a randomized result, and both differences were statistically clear. They are still worth knowing when you weigh the size of the gap.

What neither trial answers

Neither tested a compounded preparation. Both ran on the approved products, made to an approved specification and supplied for the trial. Nothing here transfers automatically to a vial from a compounding pharmacy, and the legal status page covers why those are a different product rather than a cheaper version of the same one.

Neither costed anything. A trial reports what a drug does, not what it is sold for. What each seller charges for semaglutide, restated as a true calendar month, is on the index.

Neither chooses for you. Which drug suits a particular person depends on their history, what they tolerate and what a prescriber judges appropriate, and for some people tolerability matters more than the average. This page is not medical advice.