What the Trials Found on Semaglutide and Alcohol Cravings

Semaglutide, the drug in Ozempic and Wegovy, lowered alcohol craving in randomized trials. In a 9-week trial of 48 adults, craving and drinks per drinking day fell but drinking days did not (Hendershot et al., 2025). In a 26-week trial of 108 people with obesity, heavy drinking days fell 13.7 percentage points more than on placebo (Klausen et al., 2026).

Key facts

• A third trial, 8 weeks of semaglutide tablets in 50 adults seeking treatment, missed its primary outcome (Schacht et al., 2026).
• No US or UK GLP‑1 label we checked on September 24, 2026 lists alcohol use disorder among its uses (Wegovy label, section 1).
• Three medicines are approved for alcohol use disorder in the US: naltrexone, acamprosate and disulfiram (NIAAA).
• For tirzepatide we found no published trial results; a phase 2 study planning to enroll 42 adults is listed as recruiting (NCT06994338).

What the trials found

We found three placebo-controlled trials of semaglutide in alcohol use disorder, published in 2025 and 2026 (Hendershot et al., 2025; Schacht et al., 2026; Klausen et al., 2026).

The 9-week trial enrolled people who were not seeking treatment, and its main measure was a lab drinking session after they had reached 0.5 mg (Hendershot et al., 2025). Drinks per drinking day and weekly craving fell, and heavy drinking fell more over time than on placebo. Average drinks per day and the number of drinking days did not change.

In the oral trial, lab craving and drinks per day did not differ from placebo (Schacht et al., 2026). Heavy drinking days, drinks per drinking day and day-to-day craving did fall. Its 3 mg and 7 mg tablets are Rybelsus strengths, and only 0.4% to 1% of a Rybelsus dose is absorbed, against 89% of a Wegovy injection (Rybelsus label, section 12.3; Wegovy label, section 12.3).

In the Lancet trial, heavy drinking days fell by 41.1 percentage points from baseline on semaglutide and 26.4 on placebo, which the trial’s statistical model put at 13.7 points apart (Klausen et al., 2026). Side effects were mostly short-lived, mild to moderate stomach symptoms, and they were more common on semaglutide.

Can you drink alcohol on Ozempic or Mounjaro?

None of the US labels for Ozempic, Wegovy, Mounjaro or Zepbound mentions drinking alcohol. The cautions that apply come from other parts of the labels, on low blood sugar and fluid loss, and from diabetes guidance.

Low blood sugar is a risk if you also take insulin or a sulfonylurea. The Ozempic label warns that this combination raises the risk of hypoglycemia, including severe hypoglycemia (Ozempic label, section 5.5). The American Diabetes Association advises monitoring glucose after drinking because of delayed hypoglycemia, "especially when using insulin or insulin secretagogues" (Standards of Care in Diabetes 2026, recommendation 5.19).

Acute kidney injury has been reported on semaglutide, mostly in people who became dehydrated from nausea, vomiting or diarrhea (Ozempic label, section 5.6). The Zepbound label tells patients to take precautions to avoid fluid depletion (Zepbound label, section 17), and the days side effects peak after a shot are the days to watch your fluids most.

In one small crossover trial in older men, 30 g of alcohol as wine or spirits raised urine output over the next 4 hours but not over 24 hours (Polhuis et al., 2017). Beer made no difference. For a daily fluid target based on your body weight, use the GLP‑1 water intake calculator.

Where to get help

If you drink heavily or every day, do not stop suddenly on your own. NIAAA describes withdrawal after long, heavy drinking as "painful or even potentially life-threatening" (NIAAA). If a GLP‑1 has already cut your drinking and you now want to stop completely, ask a doctor how to do it safely.

• In the US, SAMHSA’s National Helpline, 1-800-662-HELP (4357), gives treatment referral and information 24 hours a day (FindTreatment.gov).
• In the UK, Drinkline is free on 0300 123 1110, weekdays 9am to 8pm and weekends 11am to 4pm (NHS, alcohol support).
• A GP or primary care doctor can assess withdrawal risk and discuss approved medicines.

Is semaglutide approved for alcohol use disorder?

No. Approved uses are listed in section 1 of each US label and in section 4.1 of each UK Summary of Product Characteristics (SmPC). None includes alcohol use disorder, so prescribing one for drinking would be off-label.

Alcohol use disorder already has approved treatments: naltrexone, acamprosate and disulfiram in the US (NIAAA). In the UK, NICE recommends acamprosate, oral naltrexone and, for some people, disulfiram alongside psychological treatment (NICE CG115).

Does Mounjaro reduce alcohol cravings?

We found no published randomized trial of tirzepatide in alcohol use disorder. A phase 2 trial at the University of Southern California plans to enroll 42 adults with alcohol use disorder and overweight or obesity, and was recruiting at its last update (ClinicalTrials.gov NCT06994338).

In rodents, tirzepatide reduced voluntary drinking and prevented binge and relapse-like drinking (Edvardsson et al., 2026). In a remote study of 153 people with obesity who drank, those taking semaglutide or tirzepatide reported drinking less than before they started and less than a comparison group (Quddos et al., 2023). That study relied on self-report from people who had chosen their own treatment.

Why a weight-loss drug might affect drinking

In rats, semaglutide reduced alcohol intake and relapse-like drinking, and semaglutide tagged with a fluorescent marker was found in the nucleus accumbens, a reward area of the brain (Aranäs et al., 2023). In mice it blunted the rise in dopamine that alcohol causes there. Brain recordings in one person on tirzepatide linked food preoccupation to activity in the same area.

Food cravings on semaglutide can wane with time. By week 104 of STEP 5, the UK Wegovy label reports "no clear effect" on craving for sweet food (Wegovy SmPC, section 5.1), one of the patterns behind food noise coming back on a GLP‑1.

The limits of this evidence

Across the three trials, 206 people received semaglutide or placebo (48 + 50 + 108).

• The two shorter trials lasted 8 and 9 weeks, too short to show whether the effect lasts.
• The 9-week trial enrolled people who were not trying to cut down, a different group from people seeking help.
• The Lancet trial enrolled only people with obesity, and everyone also received cognitive behavioral therapy.
• The oral trial’s main measure did not improve.