Tirzepatide vs Semaglutide: differences, evidence and safety

Answer first

Semaglutide switches on the receptor for one gut hormone, GLP-1. Tirzepatide switches on two, GIP and GLP-1. Both are approved prescription medicines in eight markets. Two open-label trials compared them directly, SURPASS-2 on long-term blood sugar and SURMOUNT-5 on weight. Heart, kidney and liver results come from separate trials and cannot be ranked.

At a glance

How they work

Both drugs are peptides made in the laboratory, and both copy hormones that the gut releases after a meal ^[1], ^[2]. They differ in which hormone they are built on and in how many hormone docking points, or receptors, they switch on.

Tirzepatide

• Two receptors, GIP and GLP-1. Its chain of 39 amino acid building blocks copies the gut hormone GIP (glucose-dependent insulinotropic polypeptide). It also switches on the receptor for GLP-1 ^[1].
• Stomach and appetite. Switching on the GLP-1 receptor slows the emptying of the stomach and makes people eat less ^[1].
• Insulin and glucagon. According to the label, the body releases more insulin and less glucagon, the hormone that pushes blood sugar up. Both effects only come into play when blood sugar is high ^[1].
• The GIP share is open. The label names no trial designed to tell the two hormone systems apart ^[1].

Semaglutide

• One receptor, GLP-1. Its chain of 31 building blocks is a slightly altered copy of human GLP-1, matching it in 94% of its sequence. It leaves the receptors of related gut hormones alone ^[2].
• Insulin and glucagon. The body releases more insulin and less glucagon, and it does both only when blood sugar is actually high ^[2].
• Stomach. The package insert describes a small delay in stomach emptying in the first phase after a meal ^[2].

The core difference is the GIP receptor. Semaglutide is built on the frame of GLP-1 itself and acts on that receptor alone. Tirzepatide is built on the GIP template and acts on both ^[1], ^[2].

Both carry a fatty acid chain that makes them cling to a carrier protein in the blood. The body takes about five days to clear half a dose of tirzepatide ^[1]. Semaglutide lasts about a week and is still detectable roughly 5 weeks after the last dose ^[2].

What the studies show

Two trials that compared tirzepatide and semaglutide directly could be found. Both were open-label, meaning participants and investigators knew which drug was being given. That can colour how results are judged ^[3], ^[4].

Head-to-head: long-term blood sugar

Head-to-head, phase 3 RCT, open-label SURPASS-2 enrolled 1,879 adults with type 2 diabetes who were already taking metformin. The endpoint was glycated haemoglobin, a blood value that reflects average blood sugar over the past two to three months ^[3].

After 40 weeks it fell by 2.01, 2.24 and 2.30 percentage points across the three tirzepatide doses, and by 1.86 points on semaglutide. All three differences were large enough to count as statistically reliable. Participants on tirzepatide also lost 1.9 kg, 3.6 kg and 5.5 kg more weight than those on semaglutide ^[3].

Head-to-head: weight

Head-to-head, phase 3b RCT, open-label SURMOUNT-5 enrolled 751 adults with obesity and without diabetes. At week 72 average weight had fallen by 20.2% on tirzepatide and by 13.7% on semaglutide. Waistlines shrank by 18.4 cm against 13.0 cm ^[4].

For heart, kidney, liver and other outcomes, no head-to-head trial could be found in the sources reviewed for this page. Everything below comes from separate trials. Their people, comparisons and durations differ, so the figures cannot be set against each other.

Weight in separate trials

Phase 3 RCT SURMOUNT-1 allocated 2,539 adults with obesity and without diabetes at random to tirzepatide or a dummy injection for 72 weeks. Weight fell by 15.0%, 19.5% and 20.9% across three doses, against 3.1% on placebo ^[5].

Phase 3 RCT STEP 1 assigned 1,961 adults with obesity and without diabetes to semaglutide or a dummy treatment for 68 weeks. They lost 14.9% of their weight, against 2.4% ^[6].

Phase 3 RCT In adults who also had type 2 diabetes, both effects were smaller. SURMOUNT-2 in 938 adults recorded losses of 12.8% and 14.7% at week 72 ^[7]. STEP 2 in 1,210 adults recorded 9.6% against 3.4% on placebo ^[8].

Keeping the weight off

Phase 3 RCT, randomised withdrawal SURMOUNT-4 gave all 670 participants tirzepatide openly for 36 weeks, then split them at random. Between week 36 and week 88 those who carried on lost a further 5.5%, while those switched to placebo regained 14.0% ^[9].

Phase 3 RCT, withdrawal design STEP 4 gave semaglutide openly for 20 weeks, then split 803 adults at random. Between week 20 and week 68, those who carried on lost a further 7.9%, while those switched to placebo put 6.9% back on ^[10].

Heart

Phase 3 RCT SELECT followed 17,604 adults aged 45 and over for a median of 39.8 months. All had obesity and existing heart disease but no diabetes. A serious heart event struck 6.5% on semaglutide and 8.0% on placebo (hazard ratio 0.80, range 0.72 to 0.90) ^[11].

Phase 3 RCT, built only to show it was no worse SUSTAIN-6 followed 3,297 adults with type 2 diabetes and a high risk of heart problems over 104 weeks. A serious heart event struck 6.6% on semaglutide and 8.9% on placebo (hazard ratio 0.74, range 0.58 to 0.95) ^[12].

Phase 3 RCT, active comparator SURPASS-CVOT followed 13,299 adults with type 2 diabetes and hardened arteries for a median of 210.1 weeks. A serious heart event occurred in 12.2% on tirzepatide and 13.1% on the older drug dulaglutide (hazard ratio 0.92, range 0.83 to 1.01) ^[13].

That trial showed tirzepatide to be no worse than dulaglutide (p = 0.003). Being better was tested and not shown (p = 0.09) ^[13]. The semaglutide trials compared against a dummy treatment, SURPASS-CVOT against an active drug, so the hazard ratios answer different questions.

Heart failure

Phase 3 RCT SUMMIT followed 731 people with obesity and the stiff form of heart failure for a median of 104 weeks. Death from a heart cause or worsening heart failure occurred in 9.9% on tirzepatide and 15.3% on placebo (hazard ratio 0.62, range 0.41 to 0.95) ^[14].

Phase 3 RCT Two semaglutide trials enrolled 1,146 adults with obesity and the same form of heart failure, each over 52 weeks. Among those without diabetes, a symptom score improved by 16.6 points against 8.7. The distance walked in six minutes grew by 21.5 metres against 1.2 metres ^[15].

The semaglutide trials measured symptoms and walking, not hospital stays or deaths, so the endpoints differ from SUMMIT ^[15], ^[16].

Liver

Phase 2 RCT SYNERGY-NASH treated 190 adults for 52 weeks whose fatty liver had become inflamed and scarred, confirmed by a tissue sample. Among 157 with an assessable follow-up sample, the inflammation cleared without worse scarring in 44%, 56% and 62% on tirzepatide, against 10% on placebo ^[17].

Phase 3 RCT, interim results ESSENCE reported week 72 for its first 800 participants with the same kind of liver disease. The inflammation cleared without worse scarring in 62.9% on semaglutide against 34.3%. The scarring improved without worse inflammation in 36.8% against 22.4% ^[18].

The tirzepatide result rests on one mid-stage trial ^[17]. The semaglutide result is an interim look at a trial that runs on to 240 weeks ^[18].

Outcomes studied for only one of the two

• Semaglutide and the kidneys. FLOW enrolled 3,533 adults with type 2 diabetes and long-standing kidney disease, followed for a median of 3.4 years. The combined main outcome, kidney failure and death counted together, occurred 24% less often (hazard ratio 0.76, range 0.66 to 0.88) ^[19].
• Tirzepatide and sleep apnoea. SURMOUNT-OSA enrolled 469 adults with obesity and moderate to severe obstructive sleep apnoea. By week 52 breathing interruptions fell by 20.0 and 23.8 events per hour of sleep more than on placebo ^[20].

Endpoints that were missed

• Semaglutide in early Alzheimer's disease. Two trials in 3,808 people aged 55 to 85 failed on their main question about thinking and memory. The differences from placebo were −0.08 (p = 0.57) and 0.10 (p = 0.46), and both trials were stopped ^[21].
• Further questions in the tablet heart trial. SOUL, in 9,650 adults with type 2 diabetes, met its main goal (hazard ratio 0.86). The confirming questions, including a combined kidney measure, showed no clear difference ^[22].

Safety

The main side-effect figures for each drug come from separate trial pools in the US product information, gathered in different people at different doses. The lists below are therefore not a ranking. In SURPASS-2, where both were given in the same trial, nausea affected 17% to 22% on tirzepatide and 18% on semaglutide ^[3].

Tirzepatide. The pooled weight-loss trials covered 3,477 people on the drug ^[23].

• Stomach and bowel. Nausea affected 25–29% against 8% on placebo, diarrhoea 19–23% against 8%, vomiting 8–13% against 2% and constipation 11–17% against 5% ^[23].
• Stopping treatment. Side effects led 4.8–6.7% of participants to stop, against 3.4% on placebo ^[23].
• Label warnings. Pancreas inflammation, kidney damage from fluid loss, gallbladder trouble, low blood sugar alongside insulin and worsening diabetic eye disease ^[1], ^[23].

Semaglutide. The pooled weight trials covered 2,116 people on the drug and 1,261 on a dummy treatment ^[24].

• Stomach and bowel. Nausea affected 44% against 16%, diarrhoea 30% against 16%, vomiting 24% against 6% and constipation 24% against 11%. Severe reactions of this kind occurred in 4.1% against 0.9% ^[24].
• Gallbladder. Gallstones occurred in 1.6% against 0.7% ^[24].
• Eyes. SUSTAIN-6 recorded complications of diabetic damage to the retina in 3.0% against 1.8% (hazard ratio 1.76) ^[12].
• Sudden loss of vision. In June 2025 the European committee added optic nerve damage (non-arteritic anterior ischaemic optic neuropathy) as a very rare side effect, affecting up to 1 in 10,000 people treated ^[25]. The British regulator followed in February 2026 ^[26]. On 6 September 2026 the American package inserts carried no such entry ^[24].
• Broken bones. In SELECT, women broke a hip or pelvis in 1.0% against 0.2% ^[24].

Warnings on both labels.

• Boxed warning about thyroid tumours in rodents. Tirzepatide causes tumours in the C-cells of the thyroid gland in rats, and semaglutide in rats and mice. They occur more often at higher doses and longer exposure. Whether this applies to people is unknown ^[1], ^[24].
• Who must not take them. People with medullary thyroid cancer, personally or in the family, or with the inherited condition multiple endocrine neoplasia type 2 ^[1], ^[24].
• Pancreas inflammation. On 29 January 2026 the British regulator strengthened this warning for GLP-1 and dual GLP-1/GIP drugs. It pointed to cases in which tissue died and cases that ended in death ^[27].

Why this page lists no doses. Tirzepatide and semaglutide are prescription-only wherever they are approved. Choosing an amount is a matter for the product information and the prescribing doctor. This page therefore names doses only by their rank among those tested, never as an amount.

Legal status

Both are approved, prescription-only medicines in eight markets, including the United States, the European Union, the United Kingdom and Japan. Semaglutide came first: the United States approved it on 5 December 2017 and the European Union on 8 February 2018 ^[28], ^[29]. Tirzepatide followed on 13 May 2022 and 15 September 2022 ^[30], ^[31]. Approved uses differ by drug and market. In sport, neither is on the 2026 prohibited list ^[32], ^[33]. Both sit in section 6 of the penalty-free monitoring programme, semaglutide since 2024 and tirzepatide since 1 January 2026 ^[34], ^[35]. Both monographs reflect the position on 6 September 2026.

Used together?

The myPeptides blend directory lists no product that combines tirzepatide and semaglutide. Its only entry containing semaglutide, Cagrilintide + Semaglutide, pairs it with cagrilintide, a copy of the appetite hormone amylin. The directory notes that this pairing has been studied as a fixed combination in trials and is not approved.

No study of tirzepatide and semaglutide given together could be found in the sources reviewed for this page. SURPASS-2 and SURMOUNT-5 involve both, but they pit one against the other rather than combining them ^[3], ^[4]. Both drugs already act on the GLP-1 receptor ^[1], ^[2].

Frequently asked questions

Which is better, tirzepatide or semaglutide?

In SURMOUNT-5, among 751 adults with obesity and without diabetes, weight fell by 20.2% on tirzepatide and 13.7% on semaglutide at week 72. In SURPASS-2, among 1,879 adults with type 2 diabetes, long-term blood sugar fell further on all three tirzepatide doses than on semaglutide. Both trials were open-label. No head-to-head trial of heart, kidney or liver outcomes could be found.

Is tirzepatide or semaglutide safer?

The sources do not rank them. In SURPASS-2, among 1,879 adults with type 2 diabetes, nausea affected 17% to 22% on tirzepatide and 18% on semaglutide. Outside that trial, the figures come from separate trial pools. With both drugs, nausea, diarrhoea, vomiting and constipation are the most common complaints, and both labels warn about pancreas inflammation and gallbladder trouble.

What is the difference between tirzepatide and semaglutide?

Semaglutide copies one gut hormone, GLP-1, and switches on its receptor alone. Tirzepatide switches on two receptors, GIP and GLP-1, from a molecule built on the GIP template. Semaglutide exists as an injection and as a tablet, while tirzepatide is injected. Both are approved prescription medicines in eight markets.

Can tirzepatide and semaglutide be taken together?

No study of the two given together could be found in the sources reviewed for this page. The myPeptides blend directory lists no combination of them. SURPASS-2 and SURMOUNT-5 compare one against the other. Both drugs act on the GLP-1 receptor, and both are available only on prescription.

Has tirzepatide been tested directly against semaglutide?

Yes. Two such trials could be found, and in both everyone knew which drug was being given. SURPASS-2 compared them on long-term blood sugar in 1,879 adults with type 2 diabetes over 40 weeks. SURMOUNT-5 compared them on weight in 751 adults with obesity and without diabetes over 72 weeks.

What did the heart trials of tirzepatide and semaglutide show?

In SELECT, 17,604 adults had obesity and heart disease but no diabetes. Serious heart events were a fifth less frequent on semaglutide than on placebo (hazard ratio 0.80). In SURPASS-CVOT, among 13,299 adults with type 2 diabetes, tirzepatide proved no worse than the older drug dulaglutide, but being better was not shown. No head-to-head heart trial could be found.

Are tirzepatide and semaglutide both injections?

Tirzepatide is injected under the skin. Semaglutide comes as an injection under the skin and as a tablet, which contains an extra ingredient that helps the peptide cross the stomach wall. In OASIS 1, among 667 adults without diabetes, the tablet lowered weight by 15.1% over 68 weeks, against 2.4% on placebo.

Are tirzepatide and semaglutide banned in sport?

No. Neither appears on the 2026 World Anti-Doping Agency prohibited list, in or out of competition. Both are named in section 6 of the 2026 monitoring programme, semaglutide since 2024 and tirzepatide since 1 January 2026. Being monitored carries no penalty.

How long do tirzepatide and semaglutide stay in the body?

The body takes about five days to clear half a dose of tirzepatide. Semaglutide lasts about a week and can still be detected roughly 5 weeks after the last dose. In both, a fatty acid chain makes the peptide cling to a carrier protein in the blood, which slows its removal.