On the tirzepatide weight label, hypersensitivity reactions are a common adverse reaction with a printed rate: 5% at each of the three maintenance doses against 3% on placebo, sitting in the same table as nausea and vomiting.[1] On the semaglutide label, rash and urticaria carry no adult rate at all — they appear in the postmarketing list, where the label states frequency cannot be reliably estimated, and in a pediatric table.[2] Meanwhile the largest reporting-database analysis finds skin events less frequently reported on this class than on DPP-4 inhibitors, at a proportional reporting ratio of 0.27 (95% CI, 0.257 to 0.284).[3] Three instruments, three different answers about which drug is worse for the skin, and the one that reads most alarming is the only one with a denominator.
What the tirzepatide label actually measured
The 5% figure is a composite worth unpacking, because the label breaks it into two timing windows. In the pooled weight-reduction trials, immediate hypersensitivity reactions — within one day of a dose — occurred in 2.1% of treated patients against 0.4% on placebo, while non-immediate reactions occurred in 3.5% against 2.7%.[1] The label adds that the majority of these were skin reactions such as rash and itching.
Read the two windows separately and the finding sharpens. The placebo-adjusted excess is 1.7 points in the immediate window and 0.8 points in the delayed one, so most of what the drug adds arrives within a day of an injection. The diabetes label reports the same category differently, at 3.2% against 1.7%, describing them as sometimes severe and naming urticaria and eczema as examples.[4]
Both tirzepatide labels then tie the effect to immune response rather than to chance. Hypersensitivity reactions occurred in 6.2% of patients who developed anti-tirzepatide antibodies against 3% of those who did not in the weight trials, and in 106 of 2,570 (4.1%) against 73 of 2,455 (3.0%) across seven pooled adult trials.[1][4] The same antibody split governs the local reactions covered in the injection-site article, which is a different event at a different place on the body and should not be read as this one.
The semaglutide label has no adult number, and one pediatric one
Nothing in the adult semaglutide weight-reduction tables reports rash or urticaria, because neither reached the 2% threshold while exceeding placebo that the table requires. What the label does print comes from the pediatric trial: among 133 adolescents treated against 67 on placebo, rash was reported by 3% against 0% and urticaria by 3% against 0%.[2] The label states directly that pediatric patients aged 12 and older had greater incidences of rash and urticaria than adults on the same product.[2] That trial randomized 201 adolescents in a 2:1 ratio over 68 weeks, so each of those percentages rests on roughly four events in a group of 133 — small enough that the direction carries more than the magnitude.[5] Its wider results are in the adolescent article.
Liraglutide is the one molecule with an adult allergy figure, and it is small. Urticaria was reported in 0.7% of treated patients against 0.5% on placebo — a two-tenths-of-a-point gap across a program of several thousand people.[6] The same paragraph lists anaphylactic reactions, asthma, bronchospasm, oropharyngeal swelling, facial swelling, angioedema and type IV hypersensitivity reactions as having occurred in trials, without rates.
So the labeled picture is narrow and lopsided. One molecule carries a measured 5% against 3%. One carries a measured 0.7% against 0.5%. One carries nothing in adults and 3% against 0% in adolescents. Anyone weighing the two molecules most buyers choose between on this particular effect is comparing a number against an absence, which is not a comparison.
The contraindication is the part that is identical everywhere
Every label in this class — both semaglutide products, both tirzepatide products and liraglutide — is contraindicated in patients with a prior serious hypersensitivity reaction to the drug or to any of its excipients, and each warns that anaphylaxis and angioedema have been reported postmarketing.[1][2][4][6][7] Each instructs discontinuing the drug if a hypersensitivity reaction occurs and seeking medical attention promptly.
One sentence repeats across all of them and is worth quoting for what it concedes: anaphylaxis and angioedema have been reported with other GLP-1 receptor agonists, and caution is advised in a patient with such a history on another agent in the class because it is unknown whether such patients will be predisposed to the same reaction on this one.[1][2] That is the manufacturers stating plainly that cross-reactivity within the class has not been established in either direction. Switching molecules after a serious reaction is not a documented workaround, and it is not the sort of decision a checkout form is equipped to make — the wider list of people the labels rule out is in the contraindication article.
Severe reactions have a rate on exactly one label. In the pooled tirzepatide weight trials, 0.1% of treated patients had severe hypersensitivity reactions against no placebo patients.[1] One in a thousand, against zero, is the only randomized estimate of the serious end of this that exists.
What the reporting databases add, and what they invert
A 2026 analysis of six years of FDA adverse event reports coded rash, pruritus, urticaria, alopecia and angioedema across semaglutide, liraglutide, exenatide and dulaglutide. Cutaneous events appeared in up to 8.16% of cases, more often in women, at a mean age of 60. Semaglutide carried the highest raw rate and dulaglutide the lowest. Then the comparator arrives: against DPP-4 inhibitors the proportional reporting ratio was 0.27 (95% CI, 0.257 to 0.284), and in the regression exenatide carried the increased odds (OR 5.01; 95% CI, 4.69 to 5.35) while liraglutide and semaglutide showed decreased odds.[3]
Two of those results contradict each other on first reading and do not on second. Semaglutide generates the most skin reports because it generates the most reports; adjusted against a reference drug in the same class it falls below the line. The agent that rises is exenatide, an older extended-release depot — the same molecule that carries this class’s largest local reaction rate.
A separate retrospective analysis of the same database ranked the five most common cutaneous reactions as eczematous reactions, pruritus, drug eruptions, hyperhidrosis and alopecia, and put life-threatening cutaneous events at 2.17% of all cutaneous reactions, with no statistically significant difference between drug types. It also found that use for type 2 diabetes carried significantly higher rates of alopecia and hyperhidrosis than use for weight management (P = 0.000 for both).[8] Neither analysis can produce an incidence: a spontaneous reporting database has no denominator, is shaped by launch timing and marketing volume, and counts a report rather than a patient.
Beyond rash and hives, the literature is a shelf of single cases
This is where the honest answer is a census rather than a summary. A 2024 review of rare cutaneous reactions across the class found the published record consists of case reports and small series describing dermal hypersensitivity reactions, eosinophilic panniculitis, bullous pemphigoid and morbilliform drug eruptions, and states that little is known about GLP-1-induced cutaneous reactions as a category.[9] A separate 2024 paper announces itself as the first report of leukocytoclastic vasculitis induced by once-weekly semaglutide.[10] And the only published protocol for continuing treatment after a documented allergy to a drug in this class is a single exenatide desensitization case from 2023.[11]
One case series, one first report, one desensitization protocol. That is the entire indexed literature on the severe end of this, against tens of millions of prescriptions. It is not evidence that severe reactions do not happen; it is evidence that nobody has assembled enough of them to characterize who they happen to. Any page offering a risk profile for bullous pemphigoid on semaglutide is working from something other than this evidence.
What a compounded product adds that an approved one does not
Every percentage above was produced under randomization, with a known molecule at a known dose in a formulation whose complete excipient list is printed on a label. Most sellers covered here dispense compounded semaglutide or tirzepatide, which is not FDA-approved and is not reviewed by the FDA for safety, efficacy or quality before it is dispensed.
For a hypersensitivity question specifically, the gap is not abstract. A 2026 survey of 33 compounded semaglutide and tirzepatide products advertised as differing from the approved ones found that 48% combined the active drug with some mixture of cyanocobalamin, glycine, niacinamide, docusate or ondansetron, and that only 18% published a beyond-use date alongside preferred storage conditions. The authors found little justification for adding nutrients or docusate sodium to a subcutaneous product.[12] A contraindication written as “prior serious hypersensitivity to the drug or any of its excipients” cannot be applied to a vial whose excipients are not published, and a reaction to a blend cannot be attributed without knowing what was in it. What a label-grade ingredient list looks like is in the vial article.
The preservatives in multi-dose presentations are a separate matter and cut across brand and compounded alike: benzyl alcohol, which appears in the tirzepatide multi-dose vial, was named Allergen of the Year for 2026 by the American Contact Dermatitis Society on the grounds that it causes allergic contact dermatitis with repeated exposure and is missing from many standard patch-test series.[13] That is a delayed contact reaction at the skin rather than the systemic hypersensitivity this page is about, and the two are treated differently. The pharmacovigilance record for compounded products points at a third mechanism again: of 81,078 GLP-1 reports, the 707 involving compounded products carried reporting odds of 19.00 for contamination and 48.92 for preparation errors.[14] A contaminated injection can present as an inflammatory skin reaction that is not an allergy at all, which is why checking the pharmacy matters as much as checking the molecule — the vetting article sets out how.
Where the line sits
Rash and itching on this class are common enough to reach a label table on tirzepatide at 5% against 3%, uncommon enough to miss the table entirely on adult semaglutide, and mostly mild wherever they are recorded. Severe hypersensitivity has one randomized estimate anywhere: 0.1% against 0%. The full tolerability picture those sit inside is in what the trials recorded.
The presentation that is not a rash and is not a wait-and-see is the one every label in the class describes: swelling of the face, lips, tongue or throat, hives spreading over the body, difficulty breathing or swallowing, dizziness or a rapid pulse, usually within minutes to hours of a dose. That is anaphylaxis or angioedema, it is an emergency call rather than a message, and a documented serious reaction makes the drug contraindicated rather than something to retry at a lower dose.