Tirzepatide vs Semaglutide: What the SURMOUNT-5 Head-to-Head Trial Actually Shows

SURMOUNT-5 (a randomized, open-label, 72-week trial directly comparing tirzepatide and semaglutide for obesity, published in the New England Journal of Medicine in 2025) was designed to answer the question that indirect cross-trial comparisons could only estimate: when the same participants are randomized to one compound or the other, which produces greater weight reduction?

The published result was a superiority finding for tirzepatide. On the primary endpoint - percent change in body weight from baseline to week 72 - the trial reported a least-squares mean reduction of approximately 20.2% for tirzepatide versus approximately 13.7% for semaglutide. That absolute gap of roughly 6.5 percentage points corresponds to tirzepatide producing on the order of 47% greater relative weight reduction in this population.

Crucially, both arms used the maximum tolerated maintenance dose rather than a fixed dose, which removes the most common objection to head-to-head comparisons: that one drug was simply dosed more aggressively than the other. Within the limits of a single trial, the design was structured to make the comparison as fair as an open-label study allows.

The trial enrolled adults with obesity, defined by BMI (body mass index, a weight-to-height ratio calculated as kilograms divided by height in meters squared) of 30 or higher, or 27 or higher with at least one weight-related complication, and specifically excluded participants with type 2 diabetes. Roughly 751 participants were randomized 1:1 to subcutaneous tirzepatide or subcutaneous semaglutide once weekly.

Both compounds were titrated to their maximum tolerated maintenance dose over the standard escalation schedule: tirzepatide to 10 mg or 15 mg weekly, and semaglutide to 1.7 mg or 2.4 mg weekly. Treatment continued for 72 weeks. The trial was open-label, meaning participants and investigators knew which compound was assigned - a design choice driven by the different injection devices and titration schedules of the two products.

The primary endpoint was percent change in body weight at week 72. Key secondary endpoints included the proportion of participants reaching weight-reduction thresholds (5%, 10%, 15%, 20%, and 25%) and change in waist circumference. This structure mirrors the earlier SURMOUNT and STEP program endpoints, which makes the results easier to situate against the individual-drug trials that preceded it.

The headline comparison is cleanest in table form. The figures below are the published SURMOUNT-5 results at 72 weeks, shown alongside the widely cited single-drug benchmarks from SURMOUNT-1 (tirzepatide) and STEP-1 (semaglutide) for context. Note that the single-drug trials used different populations and durations, so they are reference points, not direct comparisons.

The internal consistency is notable: the head-to-head SURMOUNT-5 numbers land close to what each compound produced in its own pivotal trial. That coherence strengthens confidence that the observed difference reflects a genuine pharmacological gap rather than a quirk of one study population.

Weight is a single number that hides where the change is occurring. Waist circumference is a rough proxy for central (visceral) adiposity, which is the fat depot most tightly linked to metabolic risk in the epidemiological literature.

SURMOUNT-5 reported a greater waist circumference reduction in the tirzepatide arm, on the order of 18.4 cm versus 13.0 cm for semaglutide at 72 weeks. The direction and magnitude track the overall weight difference, which is the expected pattern - the trial did not report that one compound preferentially spared or targeted central fat in a way that diverged from total weight change.

The study did not include gold-standard body composition imaging (such as DXA or MRI) as a primary readout across the full cohort, so the fat-versus-lean-mass partitioning of the weight loss cannot be resolved from the primary published endpoints alone. That distinction matters for metabolic research and is one of several questions the trial leaves open.

A reasonable prior is that a compound producing substantially more weight loss might do so partly through harsher gastrointestinal effects, since GI adverse events (nausea, vomiting, diarrhea, and constipation, the dose-limiting effects shared across GLP-1 receptor agonists) are the classic driver of the appetite-suppression mechanism. SURMOUNT-5 provides a useful test of that assumption.

The published safety data reported that gastrointestinal events were the most common adverse events in both arms and were broadly similar in type and predominantly mild to moderate. In other words, the larger effect in the tirzepatide arm was not accompanied by a categorically worse GI burden. Both compounds share the GLP-1 receptor mechanism that produces these effects, and the standard dose-escalation schedule is designed to blunt their intensity.

This is one of the more consequential findings for interpreting the trial. If tirzepatide's advantage had come entirely at the price of dramatically higher intolerability, the comparison would read differently. Instead, the data suggest a genuine efficiency difference in the dose-response relationship rather than a simple trade of side effects for results.

The two compounds are not the same class of molecule. Semaglutide (a GLP-1 receptor agonist, marketed as Wegovy for obesity) acts through a single incretin receptor. Tirzepatide (a dual GLP-1 and GIP receptor agonist, marketed as Zepbound for obesity) activates a second receptor system as well.

The GIP receptor (glucose-dependent insulinotropic polypeptide receptor, expressed in adipocytes, pancreatic beta cells, and central nervous system regions involved in appetite and nausea signaling) is the differentiating feature. The prevailing mechanistic hypothesis is that concurrent GIP receptor agonism amplifies the metabolic and central satiety effects of GLP-1 receptor activation, and may also modulate the nausea signaling that limits how far GLP-1 monotherapy can be pushed. SURMOUNT-5 is consistent with that hypothesis: greater weight reduction without a proportionally greater GI penalty.

It is worth stating the limit of this reasoning. SURMOUNT-5 is an outcomes trial, not a mechanistic one. It demonstrates that the dual-agonist produced a larger effect in this population; it does not directly measure the GIP receptor's separable contribution. For a deeper treatment of how the receptors combine, see the mechanism analyses linked at the end of this article.

Honest interpretation requires naming the boundaries of a single study.

It was open-label. Participants knew their assignment. For a subjective, behavior-influenced outcome like body weight, unblinding can introduce bias, and blinding was not feasible given the different devices and schedules. The magnitude and consistency of the effect argue against unblinding explaining the whole result, but it cannot be excluded as a contributor.

It studied a specific population. Adults with obesity and without diabetes. The comparison may not extend to type 2 diabetes populations, where the separate SURPASS-2 trial examined the two compounds and reported its own results, nor to other research contexts.

It measured 72 weeks, not durability. The trial does not establish what happens on continued treatment beyond that window, nor what happens after discontinuation - a question the broader semaglutide literature suggests is significant, given documented weight regain after stopping.

It was not a cardiovascular outcomes trial. SURMOUNT-5 did not measure hard cardiovascular endpoints. Semaglutide has published cardiovascular outcome data from separate dedicated trials; tirzepatide's cardiovascular outcome program is ongoing. Weight change is a surrogate, not proof of downstream clinical benefit.

It compared two compounds, not the frontier. Neither drug represents the current experimental ceiling. Triple agonists such as retatrutide reported larger Phase 2 weight reductions still, in a separate trial that cannot be directly compared to SURMOUNT-5.

SURMOUNT-5 is best understood as one clean data point in a longer trajectory. The progression runs from single GLP-1 agonism (semaglutide) to dual GLP-1/GIP agonism (tirzepatide) to triple GLP-1/GIP/glucagon agonism (retatrutide, LY3437943, an investigational triple incretin receptor agonist). Each added receptor has, in its respective trial, been associated with a larger reported effect magnitude.

The head-to-head design of SURMOUNT-5 is exactly what the field has lacked elsewhere. It isolates the practical difference between adding the GIP receptor to a GLP-1 backbone, in a way that indirect comparison never could. That makes it a template: the most rigorous way to rank these compounds is to randomize them against each other, not to line up their separate trials.

For researchers studying incretin pharmacology, the implication is that cross-trial rankings should be treated as hypotheses until a direct comparison exists. Retatrutide's Phase 2 numbers are provocative, but the only way to know how a triple agonist compares to tirzepatide is a SURMOUNT-5-style head-to-head - which, as of this writing, the published literature does not yet provide.

Researchers studying GLP-1, GIP, and glucagon receptor pharmacology and multi-agonist metabolic mechanisms can review Retatrutide product specifications at Blackwell BioLabs. All batches are verified by third party testing with HPLC purity confirmation and mass spectrometry identity verification on every lot. Sold for research purposes only.

See also: Retatrutide vs Tirzepatide vs Semaglutide, Tirzepatide mechanism explained, GLP-1 mechanism explained, and Semaglutide plateau research.