Semaglutide vs Tirzepatide
Semaglutide and tirzepatide: what the trials actually compared
Two incretin analogues with one published head-to-head trial between them. What that trial measured, in whom, and the questions it was not designed to answer.
November 2, 2026 · 9 min read · 7 cited sources
A comparison limited to what was actually compared: the one trial that studied both compounds in the same participants, and the receptor-level work that explains why anyone expected a difference.
Semaglutide
- Target
- Single-target GLP-1 receptor agonist; an acylated analogue of human GLP-1 (Lau and colleagues, 2015).
- Lineage
- Designed for albumin binding and reduced enzymatic degradation, which is what extended its circulating half-life relative to native GLP-1.
- Trial programme
- Large phase 3 programmes in adults with type 2 diabetes and, separately, in adults with overweight or obesity without diabetes.
Tirzepatide
- Target
- Dual agonist at the GIP and GLP-1 receptors; characterised in vitro as imbalanced and biased between the two (Willard and colleagues, 2020).
- Lineage
- Built on a GIP-based backbone rather than a GLP-1 one, reported from discovery through clinical proof of concept by Coskun and colleagues in 2018.
- Trial programme
- Phase 3 programmes in adults with type 2 diabetes, including the one trial in which semaglutide was the active comparator.
In short
- One published trial studied both compounds in the same participants: SURPASS-2, in 1,879 adults with type 2 diabetes already taking metformin, over 40 weeks, with change in HbA1c as the primary endpoint (Frias and colleagues, 2021).
- In that trial all three tirzepatide arms showed larger mean reductions in HbA1c and in body weight than the semaglutide comparator arm. Gastrointestinal adverse events were the most frequently reported class in every arm.
- The two compounds engage different receptor sets. Semaglutide is a GLP-1 receptor agonist; tirzepatide engages the GIP receptor as well, and in-vitro work describes its activity at the two receptors as unequal rather than balanced.
- The obesity programmes — STEP 1 for semaglutide, SURMOUNT-1 for tirzepatide — are separate trials in different populations over different durations. Reading them against each other is not a head-to-head comparison, and this page does not do it.
- No published trial has compared the two compounds in a population without type 2 diabetes, and SURPASS-2 was not designed to attribute its result to the GIP component.
Side by side
Only dimensions the literature documents for both compounds appear here. Anything established for one and not the other is in the open questions below rather than in a half-filled row.
| Dimension | Semaglutide | Tirzepatide |
|---|---|---|
| Receptor targets as characterised | GLP-1 receptor (Lau 2015) | GIP and GLP-1 receptors (Coskun 2018; Willard 2020) |
| Molecular starting point | Acylated analogue of human GLP-1 | GIP-based backbone with engineered GLP-1 receptor activity |
| Role in the one shared trial | Active comparator arm, SURPASS-2 | Three ascending investigational arms, SURPASS-2 |
| Shared-trial population | 1,879 adults with type 2 diabetes on metformin, 40 weeks | Same participants, same trial, same duration |
| Shared-trial primary endpoint | Mean change in HbA1c from baseline | Mean change in HbA1c from baseline |
| Most frequently reported adverse-event class in that trial | Gastrointestinal | Gastrointestinal |
| Head-to-head evidence outside type 2 diabetes | None published | None published |
Two different receptor problems
Semaglutide and tirzepatide are both incretin analogues, which is where the resemblance stops being useful. Lau, Bloch, Schaffer and colleagues reported semaglutide's design in the Journal of Medicinal Chemistry in 2015: an analogue of human glucagon-like peptide-1 carrying an acyl chain that binds albumin, with substitutions that slow enzymatic degradation. The design problem they were solving was survival in circulation, and the target was one receptor.
Tirzepatide was not built from a GLP-1 scaffold. Coskun, Sloop, Loghin and colleagues described it in Molecular Metabolism in 2018 as a dual agonist at the receptors for glucose-dependent insulinotropic polypeptide and GLP-1, developed on a GIP-based backbone and carried through to clinical proof of concept. Two receptors, one molecule.
Willard, Douros, Gabe and colleagues then characterised that dual activity in JCI Insight in 2020, and the title of their paper is the finding: tirzepatide is an imbalanced and biased dual GIP and GLP-1 receptor agonist. In their cell-based assays its activity at the two receptors was not equivalent, and its signalling at the GLP-1 receptor was described as biased between downstream pathways rather than engaging them in the proportions native GLP-1 does. That is receptor pharmacology measured in cells. It is the honest basis for expecting the two compounds to behave differently in a trial — not a prediction of how differently, or in whom.
The one trial that compared them
Frias, Davies, Rosenstock and colleagues published SURPASS-2 in the New England Journal of Medicine in 2021. It enrolled 1,879 adults with type 2 diabetes who were already taking metformin, randomised them to one of three ascending tirzepatide arms or to semaglutide as the active comparator, and followed them for 40 weeks. The primary endpoint was mean change in glycated haemoglobin from baseline. Change in body weight was among the secondary endpoints.
What the investigators reported: every tirzepatide arm showed a larger mean reduction in HbA1c than the semaglutide arm, and the reductions increased across the ascending arms. The same ordering appeared in mean body-weight change. Gastrointestinal events — nausea, diarrhoea, vomiting — were the most frequently reported adverse-event class in all arms, and discontinuation for adverse events occurred in every arm.
This is the only published trial in which the two compounds were studied in the same participants under the same protocol, which makes it the only comparison on this page that is a comparison at all. It is also one trial, in one population, at one duration, with a glycaemic primary endpoint. A single trial establishes what it measured, not a general ranking.
The obesity programmes are not a head-to-head
The question most visitors arrive with concerns body weight, and here the literature offers no comparison at all. Wilding, Batterham, Calanna and colleagues reported STEP 1 in the New England Journal of Medicine in 2021: semaglutide in 1,961 adults with overweight or obesity and without diabetes, over 68 weeks, with percentage change in body weight as a co-primary endpoint. Jastreboff, Aronne, Ahmad and colleagues reported SURMOUNT-1 in the same journal in 2022: tirzepatide in 2,539 adults, over 72 weeks.
Different trials, different participants, different durations, different protocols, run years apart against different placebo groups. The arithmetic of putting the two reported percentages side by side is trivial, and the inference is invalid: between-trial differences in population, baseline characteristics, trial conduct and era are controlled by nothing, so a difference in reported means cannot be attributed to the compounds. Cross-trial comparison of this kind is the most common error in writing about this compound class.
If a head-to-head trial in a population without type 2 diabetes is published, the comparison becomes answerable. Until then the correct statement is that each programme measured what it measured, separately.
What SURPASS-2 could not attribute
A dual agonist outperforming a single-receptor agonist on a glycaemic endpoint invites the conclusion that the GIP receptor is responsible. SURPASS-2 cannot support that conclusion. It compared two molecules, not two mechanisms: the tirzepatide arms differ from the comparator arm in receptor coverage, molecular scaffold, binding kinetics and signalling bias all at once, and the trial contained no arm that isolates any one of those variables.
Rosenstock, Wysham, Frias and colleagues reported SURPASS-1 in The Lancet in 2021 — tirzepatide against placebo in adults with type 2 diabetes who were not taking other glucose-lowering drugs. That trial establishes an effect relative to placebo in that population. It does not dissect mechanism either, and no trial in the published programme was designed to.
The mechanistic question therefore sits where Willard and colleagues left it: in cells, where the imbalance and the signalling bias are measurable, and where their consequences for a clinical endpoint remain a hypothesis rather than a result.
On the safety record
Neither compound has a thin human safety record. Both have been studied in thousands of participants across multiple phase 3 trials with published adverse-event tables — unusual for anything else in this catalog, and worth stating for that reason.
What the comparison cannot do is rank them on safety. In SURPASS-2 gastrointestinal events were the dominant reported class in every arm, including the comparator, and the trial was powered for a glycaemic endpoint rather than for differences in uncommon events. Comparing rare events needs either much larger populations or post-marketing surveillance designed for it, and a 40-week randomised trial of this size supplies neither.
Nothing above describes what either compound would do in a person. Both are supplied here for in-vitro and laboratory research only, and no dosing, frequency or route of administration appears anywhere on this site.
What the cited work does not settle
- Whether the GIP receptor component accounts for the difference measured in SURPASS-2. That trial compared two molecules differing in several respects at once, and contained no arm isolating receptor coverage.
- How the two compounds compare in any population without type 2 diabetes. No head-to-head trial has been published, and the separate obesity trials differ in population and duration in ways no analysis of the published results can adjust for.
- Whether the biased, imbalanced receptor signalling Willard and colleagues measured in cells has any counterpart in a clinical endpoint. That link is a hypothesis, not a finding.
- How the two compare on uncommon adverse events. No published randomised comparison is powered for that question.
References
- Frias JP, Davies MJ, Rosenstock J, Perez Manghi FC, Fernandez Lando L, Bergman BK, Liu B, Cui X, Brown K. Tirzepatide versus Semaglutide Once Weekly in Patients with Type 2 Diabetes. New England Journal of Medicine, 2021; 385(6): 503-515.
- Coskun T, Sloop KW, Loghin C, Alsina-Fernandez J, Urva S, Bokvist KB, Cui X, Briere DA, Cabrera O, Roell WC, Kuchibhotla U, Moyers JS, Benson CT, Gimeno RE, D'Alessio DA, Haupt A. LY3298176, a novel dual GIP and GLP-1 receptor agonist for the treatment of type 2 diabetes mellitus: From discovery to clinical proof of concept. Molecular Metabolism, 2018; 18: 3-14.
- Willard FS, Douros JD, Gabe MB, Showalter AD, Wainscott DB, Suter TM, Capozzi ME, van der Velden WJC, Stutsman C, Cardona GR, Urva S, Emmerson PJ, Holst JJ, D'Alessio DA, Coghlan MP, Rosenkilde MM, Campbell JE, Sloop KW. Tirzepatide is an imbalanced and biased dual GIP and GLP-1 receptor agonist. JCI Insight, 2020; 5(17): e140532.
- Lau J, Bloch P, Schaffer L, Pettersson I, Spetzler J, Kofoed J, Madsen K, Knudsen LB, McGuire J, Steensgaard DB, Strauss HM, Gram DX, Knudsen SM, Nielsen FS, Thygesen P, Reedtz-Runge S, Kruse T. Discovery of the Once-Weekly Glucagon-Like Peptide-1 (GLP-1) Analogue Semaglutide. Journal of Medicinal Chemistry, 2015; 58(18): 7370-7380.
- Wilding JPH, Batterham RL, Calanna S, Davies M, Van Gaal LF, Lingvay I, McGowan BM, Rosenstock J, Tran MTD, Wadden TA, Wharton S, Yokote K, Zeuthen N, Kushner RF. Once-Weekly Semaglutide in Adults with Overweight or Obesity. New England Journal of Medicine, 2021; 384(11): 989-1002.
- Jastreboff AM, Aronne LJ, Ahmad NN, Wharton S, Connery L, Alves B, Kiyosue A, Zhang S, Liu B, Bunck MC, Stefanski A. Tirzepatide Once Weekly for the Treatment of Obesity. New England Journal of Medicine, 2022; 387(3): 205-216.
- Rosenstock J, Wysham C, Frias JP, Kaneko S, Lee CJ, Fernandez Lando L, Mao H, Cui X, Karanikas CA, Thieu VT. Efficacy and safety of a novel dual GIP and GLP-1 receptor agonist tirzepatide in patients with type 2 diabetes (SURPASS-1): a double-blind, randomised, phase 3 trial. The Lancet, 2021; 398(10295): 143-155.
What this page is, and is not
This is a summary of published research, written for qualified professionals evaluating compounds for laboratory work. Both compounds discussed are supplied by strictly for in-vitro and laboratory research use. Neither is a drug, a dietary supplement, or a cosmetic; neither is intended for human or veterinary use; and nothing above is medical advice, a treatment recommendation, or a claim that either compound produces any effect in a person. We publish no dosing or administration guidance of any kind, and no page on this site recommends one compound over another.
