Tirzepatide vs Semaglutide Side Effects: What the Head-to-Head Evidence Shows

The most common mistake in comparing these two drugs is putting the Zepbound label’s percentages next to Wegovy’s and treating the difference as meaningful.

You cannot do that, and both labels say so. Every FDA label carries a version of the same statement: adverse reaction rates observed in the clinical trials of one drug cannot be directly compared to rates in the trials of another drug, because the trials enrolled different people under different conditions [2][3]. The Zepbound numbers come from 2,519 adults in 72-week weight-reduction trials. The Wegovy numbers come from a different set of trials run at a different time. Any gap between them mixes real drug differences with trial design differences, and there is no way to separate the two after the fact.

What you can use is a trial that gave both drugs to similar people at the same time under the same rules. There is exactly one of those, and there is one high-quality observational study that comes close. This page is built on those.

A quick naming note. Tirzepatide is sold in the US as Mounjaro (type 2 diabetes and cardiovascular risk) and Zepbound (weight management and obstructive sleep apnea). Semaglutide is sold as Ozempic and Rybelsus (type 2 diabetes) and Wegovy (weight management). Tirzepatide acts on two gut hormone receptors, GIP and GLP-1; semaglutide acts on one [2][3].

Nothing here is medical advice.

What did the head-to-head trial find?

SURMOUNT-5 is the only randomized trial that put the two drugs against each other. It enrolled 751 adults with obesity but without type 2 diabetes, ran 72 weeks, and compared the maximum tolerated dose of tirzepatide (10 or 15 mg) against the maximum tolerated dose of semaglutide (1.7 or 2.4 mg) [1].

Here are the safety results side by side [1]:

Outcome	Tirzepatide	Semaglutide
Any adverse event	76.7%	79.0%
Serious adverse events	4.8%	3.5%
Deaths	0	0
Stopped treatment for any adverse event	6.1%	8.0%
Stopped treatment for a gut adverse event	2.7%	5.6%
Nausea	43.6%	44.4%
Constipation	27.0%	28.5%
Diarrhea	23.5%	23.4%
Vomiting	15.0%	21.3%
Gastroesophageal reflux	6.1%	10.6%
Dyspepsia	5.9%	7.4%
Dizziness	6.4%	4.8%
Fatigue	10.4%	12.2%
Hair loss (alopecia)	8.3%	6.1%
Decreased appetite	4.5%	5.1%
Injection site reaction	8.6%	0.3%

Read that table carefully, because the headline and the detail point different ways.

The headline: overall side effect burden was similar, and slightly more people stopped semaglutide than tirzepatide.

The detail: the two drugs are not identical in which side effects they produce.

Where do the two drugs genuinely differ?

Four places stand out.

Vomiting and reflux favor tirzepatide

Vomiting was about a third less common on tirzepatide (15.0% versus 21.3%), and reflux was nearly half as common (6.1% versus 10.6%) [1]. Nausea, constipation and diarrhea were effectively tied.

This is the most practically useful finding in the trial, because vomiting is the symptom most likely to dehydrate someone, and dehydration is the pathway both labels describe to acute kidney injury [2].

Gut-related quitting favors tirzepatide

Only 2.7% of people stopped tirzepatide because of gut side effects, against 5.6% on semaglutide [1]. That is the same direction the overall discontinuation number points, and it is consistent with the vomiting result.

Injection site reactions strongly favor semaglutide

This is the largest single difference in the trial, and it goes the other way: 8.6% on tirzepatide versus 0.3% on semaglutide [1].

The Zepbound label fills in why this matters. Injection site reactions there occurred in 6% to 8% versus 2% on placebo — but in people who developed anti-tirzepatide antibodies, the rate was 11.3%, against 1% in those without [2]. And most people on tirzepatide do develop those antibodies: 64.5% in the weight-management trials [2]. The label says the antibodies had no identified effect on how well the drug worked, only on reaction rates.

The term covers bruising, redness, itching, pain and rash at the injection spot. It is a nuisance category, not a dangerous one, but it is the single most reliable tolerability difference between the two drugs.

Hair loss slightly favors semaglutide, on weak data

8.3% versus 6.1% in SURMOUNT-5 [1]. That is a small gap in a trial that was not designed to measure hair loss and did not use dermatologist assessment. It is also the reverse of what the labels suggest at first glance: hair loss is on the Zepbound label with percentages but appears on the Wegovy label too, and the strongest population-level evidence covers the whole class rather than distinguishing the drugs.

What does the observational evidence say?

Randomized trials are short and enroll carefully selected people. For serious, rare gut outcomes, the useful comparison is a large cohort study.

The best one published so far is a new-user, active-comparator cohort study in adults with type 2 diabetes covering January 2019 to August 2024. In 46,620 propensity-matched pairs, the hazard ratio for a composite of acute pancreatitis, biliary disease, bowel obstruction, gastroparesis and severe constipation was 1.07 (95% CI 0.90 to 1.26) for tirzepatide versus subcutaneous semaglutide [4].

A hazard ratio of 1.07 with a confidence interval straddling 1.0 means: no detectable difference. The authors concluded the two agents have similar gastrointestinal safety profiles [4].

Analyses of the FDA’s spontaneous reporting database mostly agree, with a few differences at the margins. One 2026 analysis found tirzepatide had lower reporting odds than semaglutide for vomiting, constipation and decreased appetite [8]. Reporting data measure how often something gets reported, not how often it happens, so those comparisons are suggestive rather than conclusive.

Where do the two drugs really diverge on vision risk?

There is one place where the pharmacovigilance data separate the two sharply, and it is the eye.

NAION — non-arteritic anterior ischemic optic neuropathy — is sudden, usually painless vision loss in one eye caused by reduced blood flow to the optic nerve. It is rare, and it is most often permanent.

A 2026 comparative analysis of the FDA’s adverse event database across six incretin therapies through Q3 2025 found [5]:

Drug	Optic ischemic neuropathy reports	Reporting odds ratio
Semaglutide	355	94.45 (95% CI 83.02 to 107.45)
Liraglutide	13	4.58 (95% CI 2.65 to 7.91)
Tirzepatide	49	2.94 (95% CI 2.20 to 3.93)
Dulaglutide, exenatide, lixisenatide	—	no signal

One wording note: the study counted the MedDRA term “optic ischaemic neuropathy” rather than NAION itself, because NAION is not a separate coded term. Tirzepatide’s signal is small but it is statistically significant — its confidence interval does not cross 1 — so this is a thirty-fold difference in signal strength, not a signal versus no signal. The authors’ own conclusion is that the absence of a signal across the whole class argues against a uniform class effect [5].

The regulators have acted the same way. European regulators added NAION to semaglutide product information in 2025 as a very rare side effect. They did not add it to tirzepatide’s. In the US, the FDA listed NAION as a potential class signal naming both tirzepatide brands in its October–December 2024 quarterly report, but as of September 14, 2026 no NAION warning appears on any US GLP-1 label, semaglutide or tirzepatide [2][7].

Three cautions before anyone reads too much into the gap.

First, reporting-database signals are heavily influenced by publicity. A September 2026 preprint argues that the semaglutide NAION signal shows hallmarks of notoriety bias, with reporting for other drugs in the class rising only after mid-2024 media coverage. That analysis has not been peer reviewed.

Second, the large target trial emulation most often cited on this topic — 159,398 propensity-matched patients, hazard ratio 1.76 for NAION — combined semaglutide and tirzepatide into a single exposure group. It cannot tell you anything about one drug versus the other.

Third, the American Academy of Ophthalmology’s May 2026 clinical statement says the evidence suggests a possible association between semaglutide and NAION of roughly two-fold, that causality has not been established, and that the association with other GLP-1 drugs — tirzepatide specifically — has been less frequently studied and warrants further investigation [6]. “Less studied” is not the same as “safer.”

What is on one label and not the other?

Comparing the actual FDA documents is more reliable than comparing percentages, because it shows what each regulator concluded the evidence supported.

On the tirzepatide labels, not on the semaglutide labels:

• An oral contraceptive instruction. Both Mounjaro and Zepbound tell people using oral hormonal contraceptives to switch to a non-oral method, or add a barrier method, for 4 weeks after starting and for 4 weeks after each dose increase. Semaglutide labeling carries no such instruction. The pharmacokinetic data behind it: with a single 5 mg tirzepatide dose, peak levels of the contraceptive hormones fell by 55% to 66% and overall exposure by 20% to 23% [2][7].
• Hair loss with incidence figures (on Zepbound) [2].
• Hypersensitivity reactions among the most common adverse reactions (on Zepbound), an unusual listing [2].

On the semaglutide labels, not on the tirzepatide labels:

• Nothing of comparable weight. The warnings sets are otherwise the same items: thyroid C-cell boxed warning, MEN 2 and hypersensitivity contraindications, pancreatitis, gallbladder disease, kidney injury from dehydration, hypoglycemia with insulin or a secretagogue, diabetic retinopathy complications, severe gastrointestinal reactions, pulmonary aspiration under anesthesia, and never sharing a pen [2][3].

Both drug families also lost the suicidal behavior and ideation warning in February 2026, after the FDA’s January 13, 2026 request that manufacturers remove it. Mounjaro never carried one; Zepbound and Wegovy did [2][3].

What about long-term safety?

Here the comparison is uneven, because the two drugs have different long-term datasets.

For tirzepatide, the anchor is SURPASS-CVOT: 13,299 adults with type 2 diabetes and established heart disease, followed a median of four years against dulaglutide. Gut events were more common on tirzepatide than dulaglutide (42.5% versus 35.9%) and more people stopped (13.3% versus 10.2%). Adjudicated pancreatitis was 0.6% in both arms; severe hypoglycemia 0.7% in both [9].

For semaglutide, the equivalent anchors are its own cardiovascular outcome trials. Those have their own comparators and populations.

Neither of those trials compared the two drugs. Putting their results side by side would repeat exactly the error this page opened with.

The honest bottom line

Based on the one randomized head-to-head trial and the best matched cohort study:

• Overall side effect burden is similar. Neither drug is meaningfully gentler across the board.
• Tirzepatide has an edge on vomiting, reflux and gut-driven quitting.
• Semaglutide has a clear edge on injection site reactions.
• Serious gut outcomes look the same in 46,620 matched pairs.
• The NAION reporting signal is far stronger for semaglutide, though the data are reporting data and the drugs have been studied unequally.

And the limits. SURMOUNT-5 was open-label, industry-funded, ran 72 weeks, enrolled 751 people without diabetes, and was powered to detect weight differences, not rare harms. It is one trial. It should be treated as the best available evidence, not as a settled answer.

Which drug a given person tolerates better is not predictable from these tables. That conversation belongs with a healthcare provider who knows the rest of your medical history.

Sources

1. Tirzepatide versus semaglutide in obesity (SURMOUNT-5), New England Journal of Medicine
2. ZEPBOUND (tirzepatide) prescribing information — DailyMed
3. WEGOVY (semaglutide) prescribing information, FDA
4. Comparative gastrointestinal safety of tirzepatide and semaglutide, Annals of Internal Medicine 2025
5. Comparative NAION reporting across incretin therapies, Diabetes, Obesity and Metabolism 2026
6. American Academy of Ophthalmology clinical statement on GLP-1 receptor agonists and NAION
7. MOUNJARO (tirzepatide) prescribing information — DailyMed
8. Tirzepatide FAERS disproportionality analysis, 2026
9. SURPASS-CVOT, New England Journal of Medicine