Two head-to-head trials, both won by tirzepatide — and one dose asymmetry that makes the gap look bigger than it is
Tirzepatide won both trials that compared these drugs directly, and it did so while being discontinued less often for gastrointestinal side effects. The size of the win depends on the trial: at matched maximum tolerated doses in obesity it was 20.2% versus 13.7% body weight, but SURPASS-2 dosed semaglutide at 1 mg — its diabetes dose, not the 2.4 mg obesity dose — so that trial's weight gap is wider than a fair-dose comparison would produce. Semaglutide keeps three practical advantages: it is the only one with an oral formulation, it has the longer real-world safety record, and it carries approved indications tirzepatide does not.
| Attribute | Semaglutide | Tirzepatide |
|---|---|---|
| Receptor targets | GLP-1 only | GIP + GLP-1 (dual) |
| Head-to-head, obesity [1] | -13.7% body weight at 72 weeks (1.7-2.4 mg) | -20.2% body weight at 72 weeks (10-15 mg) — winner |
| Head-to-head, type 2 diabetes [2] | HbA1c -1.86%, weight -5.7 kg at 40 weeks (1 mg) | HbA1c -2.30%, weight -11.2 kg at 40 weeks (15 mg) — winner, but see the dose caveat |
| GI discontinuation in the obesity trial [1] | 5.6% | 2.7% |
| Oral formulation | Yes — Rybelsus daily tablet | No — injection only |
| Brand names | Ozempic, Wegovy, Rybelsus | Mounjaro, Zepbound |
| Route | Subcutaneous (weekly injection); Oral tablet available (Rybelsus) | Subcutaneous |
| Frequency | Once weekly (SC); Once daily (oral) | Once weekly |
| Typical dose range | SC: 0.25–2.4 mg/week titrated over 16 weeks; Oral: 3–14 mg/day | 2.5–15 mg/week, titrated every 4 weeks |
| Oral bioavailability | Low (~0.4–1% for oral Rybelsus formulation); requires fasting and SNAC absorption enhancer | Not available orally; injectable only (subcutaneous) |
| US regulatory status | FDA-approved (Ozempic, Wegovy, Rybelsus). Additional approvals: MASH with fibrosis (Aug 2025), oral 25 mg for weight management (late 2025). 135+ warning letters issued to GLP-1 compounders;… | FDA-approved: Mounjaro (T2D, adults and children 10+) and Zepbound (obesity/overweight with comorbidities, severe OSA including new sleep apnea approval). Compounding fully prohibited for 503B… |
Semaglutide mimics the GLP-1 hormone by binding to GLP-1 receptors on pancreatic beta cells (glucose-dependent), brain (hypothalamus appetite centers), stomach, and intestines. It stimulates insulin secretion, suppresses glucagon release, slows gastric emptying promoting prolonged fullness, and reduces food intake. A fatty acid chain (C18 diacid) modification enables albumin binding and resistance to DPP-4 degradation, extending its half-life to approximately 160-168 hours (~1 week), allowing once-weekly subcutaneous dosing. Injectable bioavailability is approximately 89%. Oral formulation (Rybelsus) requires administration on empty stomach 30 minutes before food and has lower bioavailability. Developed by Novo Nordisk, first FDA-approved December 5, 2017.
Tirzepatide (MW ~4813 g/mol, C225H348N48O68) simultaneously activates both GIP (glucose-dependent insulinotropic polypeptide) and GLP-1 (glucagon-like peptide-1) receptors. This dual agonism enhances glucose-dependent insulin secretion, suppresses glucagon, slows gastric emptying, increases satiety signaling, and improves insulin sensitivity. The added GIP activity appears to synergistically amplify metabolic effects beyond GLP-1 alone, contributing to superior weight reduction. A C20 fatty acid via hydrophilic linker promotes albumin binding, resulting in a half-life of approximately 5 days, ~80% subcutaneous bioavailability, and 99% albumin binding. Tirzepatide also significantly increases serum adiponectin, an adipokine linked to lipid and glucose metabolism regulation, suggesting potential cardioprotective implications.
Design: Open-label randomised phase 3b trial, 32 sites in the US and Puerto Rico Population: 751 adults with obesity (BMI >=30, or >=27 with a weight-related complication) and without type 2 diabetes Duration: 72 weeks Arms: Tirzepatide 10 or 15 mg weekly vs semaglutide 1.7 or 2.4 mg weekly, each titrated to maximum tolerated dose
Mean body-weight change -20.2% with tirzepatide vs -13.7% with semaglutide (p<0.001). Waist circumference -18.4 cm vs -13.0 cm. Gastrointestinal adverse events leading to discontinuation: 2.7% with tirzepatide vs 5.6% with semaglutide.
Design: Open-label randomised phase 3 trial, tirzepatide added to metformin Population: 1,879 adults with type 2 diabetes; mean diabetes duration 8.6 years, baseline HbA1c 8.28%, baseline weight 93.7 kg Duration: 40 weeks Arms: Tirzepatide 5, 10 or 15 mg weekly vs semaglutide 1 mg weekly
HbA1c fell 2.01% (5 mg), 2.24% (10 mg) and 2.30% (15 mg) with tirzepatide vs 1.86% with semaglutide. Body weight fell 7.6 kg, 9.3 kg and 11.2 kg vs 5.7 kg (all p<0.001). 51% of the tirzepatide 15 mg group reached HbA1c below 5.7% vs 20% on semaglutide.
Design: Retrospective propensity-score-matched (1:1) cohort using linked electronic health records and pharmacy dispensing data Population: 18,386 matched US adults with overweight or obesity (9,193 tirzepatide, 9,192 semaglutide), from 41,222 identified
| Compound | 3 months | 6 months | 12 months |
|---|---|---|---|
| Semaglutide | -3.6% | -5.8% | -8.3% |
| Tirzepatide | -5.9% | -10.1% | -15.3% |
Observational, not randomised. Matching reduces confounding but cannot eliminate it — prescribing choice, insurance coverage, adherence and dose escalation all differ outside a trial protocol. Hazard ratios favoured tirzepatide for reaching 5% (1.76, 95% CI 1.68-1.84), 10% (2.54, 95% CI 2.37-2.73) and 15% (3.24, 95% CI 2.91-3.61) weight loss. Read this as agreement in direction with the randomised trials, not as a substitute for them.
| Event | Semaglutide | Tirzepatide |
|---|---|---|
| Gastrointestinal events leading to discontinuation | 5.6% | 2.7% |
| Dominant adverse-event category | Gastrointestinal — nausea, vomiting, diarrhoea, constipation | Gastrointestinal — nausea, vomiting, diarrhoea, constipation |
| Severity and timing | Mostly mild to moderate, concentrated during dose escalation | Mostly mild to moderate, concentrated during dose escalation |
Both arms were titrated slowly, so these rates reflect a gradual escalation schedule rather than starting at the maximum dose. Only discontinuation was reported numerically in the primary publication and its summaries; per-event incidence rates for each arm sit in the supplementary appendix and are deliberately not reproduced here rather than estimated.
Most confusion about these two drugs is really confusion about branding, so it is worth settling first.
Semaglutide is the active molecule in Ozempic, Wegovy and Rybelsus. Ozempic is the type 2 diabetes injection, Wegovy the weight-management injection at a higher dose, and Rybelsus the daily oral tablet.
Tirzepatide is the active molecule in Mounjaro and Zepbound. Mounjaro is the type 2 diabetes injection and Zepbound the weight-management injection.
The practical consequence: two products can contain the same molecule at different strengths, in different formulations, and licensed for different populations. "Ozempic versus Mounjaro" and "Wegovy versus Zepbound" are therefore related to this comparison but not identical to it — they compare specific products at specific doses, whereas this page compares the molecules using the trials that tested them head to head.
The trial details sit in the evidence cards above, so the more useful thing to explain is how to read the gap, because it is routinely misquoted in both directions.
In SURMOUNT-5 the difference between arms was 6.5 percentage points of body weight: 20.2% versus 13.7%. Expressed relatively, the average reduction with tirzepatide was roughly 47% larger than with semaglutide. Both statements describe the same result. Neither means participants lost an extra 47% of their starting weight.
For a 100 kg starting weight the arithmetic is plain: about 20 kg lost on tirzepatide against about 14 kg on semaglutide, a difference of roughly 6 kg. That is the number worth carrying away, and it is smaller than "47% better" sounds and larger than "6.5 points" sounds.
Relative framing is what makes marketing copy exciting and what makes readers overestimate the practical difference. Percentage points are what the trial measured.
SURPASS-2 compared tirzepatide up to 15 mg against semaglutide at 1 mg. One milligram was semaglutide's approved type 2 diabetes dose at the time, so the trial was a fair test of the products as then licensed for diabetes — but it is not a fair test of the molecules for weight loss, because the 2.4 mg obesity dose is more than twice that.
That matters when the 11.2 kg versus 5.7 kg weight figure gets quoted as a general statement about the two drugs. It is a real result at those doses and nothing more. SURMOUNT-5, which titrated both drugs to maximum tolerated dose, is the trial to cite for a like-for-like weight comparison, and its gap is proportionally smaller.
Anyone presenting the SURPASS-2 weight numbers as the headline difference between semaglutide and tirzepatide is, intentionally or not, comparing a full dose against a partial one.
The intuitive expectation is that more weight loss costs more nausea. On the one tolerability measure the trial reported numerically, the opposite held: gastrointestinal adverse events severe enough to stop treatment occurred in 2.7% of the tirzepatide group and 5.6% of the semaglutide group.
That is discontinuation, not incidence, and the distinction matters. It does not establish that tirzepatide causes less nausea overall — per-event incidence for each arm was not reported in the primary publication's summaries. It establishes that fewer people abandoned tirzepatide because of gut side effects. Both drugs produce nausea, vomiting, diarrhoea and constipation, mostly mild to moderate and concentrated during dose escalation, and both arms were titrated slowly.
A plausible mechanistic explanation is that GIP agonism blunts the nausea signalling that GLP-1 agonism drives, but that is a hypothesis this trial was not designed to test.
The two randomised trials studied different people, and that alone shifts the weight figures.
SURMOUNT-5 excluded type 2 diabetes. SURPASS-2 enrolled only people who had it, with a mean diabetes duration of 8.6 years. Weight loss on incretin drugs is consistently smaller in people with type 2 diabetes than in people without, so the two populations cannot be pooled casually — a 40-week diabetes trial and a 72-week obesity trial are answering different questions.
The real-world cohort saw the same pattern from the other direction: participants without type 2 diabetes lost more weight than those with it, for both drugs, while tirzepatide remained ahead within each group. That consistency is useful. It means diabetes status changes the magnitude of the expected result without changing which drug produces more weight loss.
So when comparing figures across sources, check three things before treating them as contradictory: the population's diabetes status, the trial duration, and the dose of each arm.
Losing both head-to-head trials does not make semaglutide the wrong choice, and three differences have nothing to do with weight.
It is the only one of the two available as a tablet. Rybelsus is a daily oral formulation, which matters for anyone who will not inject; the trade-off is lower bioavailability and a requirement to dose on an empty stomach 30 minutes before food.
It has the longer record. Semaglutide was first approved in December 2017 against tirzepatide's 2022, and its trial programme extends into outcomes tirzepatide has not yet matched — cardiovascular risk reduction, kidney function decline in type 2 diabetes with chronic kidney disease, and non-cirrhotic MASH with moderate to advanced fibrosis. If the goal is anything beyond weight and glycaemia, the approved indication may decide the choice regardless of which drug loses more weight.
And its baseline efficacy is not in doubt: STEP 1 produced 14.9% mean weight loss against 2.4% for placebo over 68 weeks. Second place here is second in a strong field.
A 2025 meta-analysis restricted to direct comparisons pooled seven studies and 28,980 participants, finding a standardised mean difference of 0.75 (95% CI 0.52 to 0.92) favouring tirzepatide.
The confidence interval excludes zero, so the direction is consistent. But the authors reported heterogeneity above 90%, meaning the constituent studies disagreed with each other substantially — a mix of two randomised trials and five observational cohorts with different populations, durations and dose distributions. Pooling those produces a reliable answer to "which direction" and an unreliable one to "by exactly how much".
Treat the direction as settled and the magnitude as dose- and population-dependent.
For weight loss and glycaemic control, the two trials that compared them directly both favoured tirzepatide. In obesity it was 20.2% versus 13.7% body weight over 72 weeks at maximum tolerated doses, and gastrointestinal discontinuation was lower with tirzepatide (2.7% vs 5.6%). Semaglutide retains an oral option, a longer safety record, and approved indications tirzepatide lacks, so "better" depends on whether weight is the only goal.
In the only obesity trial that dosed both to maximum tolerated levels, 6.5 percentage points more body weight over 72 weeks — 20.2% versus 13.7%. Larger gaps you may see quoted usually come from SURPASS-2, which used semaglutide 1 mg rather than the 2.4 mg obesity dose.
Not on the measure that decides whether people stay on treatment. In SURMOUNT-5, gastrointestinal events leading to discontinuation were roughly half as common with tirzepatide (2.7%) as with semaglutide (5.6%), despite greater weight loss. Both cause dose-dependent nausea, vomiting, diarrhoea and constipation during escalation.
Only semaglutide. Rybelsus is a daily oral tablet; tirzepatide is injection-only. Oral semaglutide has lower bioavailability and must be taken on an empty stomach at least 30 minutes before food.
Yes. If you need an oral option, tirzepatide has none. If the goal includes cardiovascular risk reduction, slowing kidney function decline in type 2 diabetes with chronic kidney disease, or treating non-cirrhotic MASH, semaglutide has approved indications there that tirzepatide does not. It also has roughly five more years of post-approval use.
Because they draw on trials with different doses, durations and populations. SURPASS-2 ran 40 weeks in type 2 diabetes with semaglutide at 1 mg; SURMOUNT-5 ran 72 weeks in obesity without diabetes at maximum tolerated doses. A 2025 meta-analysis of direct comparisons found a consistent direction favouring tirzepatide but heterogeneity above 90%, meaning the underlying studies disagreed substantially on magnitude.
No. They are different molecules with different receptor targets. Semaglutide activates the GLP-1 receptor only. Tirzepatide activates both the GIP and GLP-1 receptors, which is why it is described as a dual agonist. They are in the same drug class and share a side-effect profile, but they are not interchangeable products.
Yes. Ozempic, Wegovy and Rybelsus all contain semaglutide. Ozempic is the type 2 diabetes injection, Wegovy the higher-dose weight-management injection, and Rybelsus the daily oral tablet. Same molecule, different strengths, formulations and approved uses.
Yes. Both contain tirzepatide. Mounjaro is licensed for type 2 diabetes and Zepbound for weight management, including obstructive sleep apnoea in adults with obesity. There is no oral tirzepatide product.
SURMOUNT-5. It titrated both drugs to the maximum dose each participant tolerated, so neither arm was handicapped. SURPASS-2 capped semaglutide at 1 mg, which was its approved diabetes dose but well below the 2.4 mg used for weight management, so its weight gap overstates the difference between the molecules.
In SURPASS-2 every tirzepatide dose beat the single semaglutide dose tested: HbA1c fell 2.01% at 5 mg, 2.24% at 10 mg and 2.30% at 15 mg against 1.86% for semaglutide 1 mg. But semaglutide was only tested at that one dose in that trial, so it is not a dose-for-dose comparison across the full range.
Yes, in direction. A propensity-matched cohort of 18,386 US adults found on-treatment weight change of -5.9%, -10.1% and -15.3% with tirzepatide at 3, 6 and 12 months, against -3.6%, -5.8% and -8.3% with semaglutide. It is observational, so it cannot rule out confounding from prescribing patterns, coverage or adherence.
In SURMOUNT-5, 20.2% versus 13.7% is a 6.5 percentage-point difference. Stated relatively, tirzepatide's average reduction was about 47% larger. Both describe the same result. For someone starting at 100 kg it means roughly 20 kg lost versus 14 kg — a difference of about 6 kg, not 47 kg.
For the expected size of the result, yes. Weight loss on both drugs is consistently smaller in people with type 2 diabetes. SURMOUNT-5 excluded diabetes and SURPASS-2 enrolled only people with it, which is one reason their figures differ. The real-world cohort found the same pattern, with tirzepatide ahead in both groups.
Whether the weight difference persists beyond 72 weeks, how the two compare on cardiovascular or kidney outcomes head to head, and whether tirzepatide's tolerability advantage holds outside slow-titration trial protocols. No head-to-head trial has been powered for outcome endpoints rather than weight and glycaemia.
Full profiles: Semaglutide, Tirzepatide.