Naltrexone for Alcohol Use Disorder

Naltrexone for alcohol at a glance

US guidelines cover moderate to severe alcohol use disorder

Alcohol use disorder (AUD) is the medical term for what people also call alcoholism, alcohol abuse or alcohol dependence.¹ Clinicians diagnose it from 11 criteria in the DSM-5. Examples are drinking more or longer than you meant to, trying and failing to cut down, and needing more alcohol to feel the same effect. Two or three criteria mean mild AUD, four or five moderate, and six or more severe.¹ In 2024, 27.9 million people in the US aged 12 and older had AUD in the past year.¹⁴

The APA and VA/DoD recommendations on medicines both cover moderate to severe AUD.^6,13 The trials behind them enrolled people with "alcohol dependence" under the older DSM-IV, which the VA/DoD treats as the same as moderate to severe AUD.¹³

Milder AUD has less guidance. NICE, in England, says naltrexone or acamprosate can be considered for harmful drinking or mild dependence when talking therapy alone has not helped, or when the person asks for medicine, alongside psychological therapy.¹⁰ For less severe AUD in the US, NIAAA points to treatment in primary care with brief counseling and FDA-approved medicines.⁵

It dulls the reward from drinking

Naltrexone blocks opioid receptors. NIAAA describes it as reducing the pleasurable effects of alcohol by interfering with the brain's own opioid signals in reward circuits.¹²

The tablet label calls the mechanism in alcoholism "not understood," with involvement of the opioid system suggested by early laboratory research.²

Naltrexone causes no sickness reaction when you drink, unlike disulfiram (Antabuse).² What drinking feels like while you take it has its own page.

In trials, the gains are modest

Naltrexone helps some people more than others. The tablet label says it "was not uniformly helpful to all patients" and puts the expected effect at "a modest improvement in the outcome of conventional treatment."²

That 11 is a number needed to treat (NNT): out of every 11 people who take naltrexone instead of a placebo, 1 more avoids going back to heavy drinking. The 95% confidence interval runs from 5 to 41. That interval is wide, so the benefit could be noticeably bigger or smaller than 1 in 11.⁴

The two big reviews disagree on whether naltrexone prevents any drinking at all. Both found that it reduces heavy drinking.^4,7 Trial-by-trial results, including the negative ones, are on does naltrexone work?.

The Vivitrol label reports a split. Its effects on drinking days, heavy drinking days and staying abstinent were seen in the 8% of trial patients who had not drunk in the week before the first shot. They "were not evident" in the 92% who were still drinking.³

It is used both to cut down and to stop

The APA recommendation applies whether your goal is to drink less or not at all.⁶ A SAMHSA advisory for prescribers adds that naltrexone suits either goal and can be continued if someone drinks again.¹⁶

Pooled trials found naltrexone slightly better than acamprosate for cutting heavy drinking and craving, and acamprosate slightly better for staying abstinent.¹⁷

The tablet trials used 50 mg every day; other schedules were not tested.² Taking it only before planned drinking, the Sinclair Method, is an off-label use with its own evidence.

Guidelines rank it in the top tier

In APA grades, the 1 means a firm recommendation: the benefits clearly outweigh the harms. The B means moderate-strength evidence.¹⁹ The APA gives naltrexone and acamprosate its top grade because they "have the best available research evidence."¹⁹

The VA/DoD says there is not enough evidence to prefer one top-tier drug over the other, so the choice should be individual. That comparison is naltrexone against topiramate. It does not put naltrexone level with acamprosate, which sits a tier lower.¹³

The monthly shot is for people who struggle with daily pills

The 2021 VA/DoD guideline says the two forms "have not been directly compared to evaluate whether the long-acting injectable formulation improves clinical outcomes." It suggests the shot when taking medicine reliably is a significant concern and the person agrees to monthly injections from a clinician. It also notes the shot must be kept refrigerated.¹³

Small head-to-head trials did exist by then, including a 2019 open-label trial of 237 adults with alcohol use disorder. A 2025 review that pooled the comparison studies found people stayed on the shot longer, with no difference in hospital stays or emergency visits.²⁰ The trial-by-trial results are in naltrexone vs Vivitrol. The shot itself is covered on Vivitrol.

Of five AUD medicines, only naltrexone tops both US guidelines

Naltrexone is the only one of five graded medicines that gets the top rating from both US guidelines: grade 1B from the APA and "strong for" from the VA/DoD. Acamprosate shares the APA grade but gets a weaker VA/DoD rating, and off-label topiramate gets the reverse.^6,13 Only naltrexone, acamprosate and disulfiram are FDA-approved for AUD.¹ All of them, plus gabapentin, baclofen and semaglutide, are compared on medication to stop drinking. If one medicine does not help, NIAAA tells clinicians that trying another is often worthwhile.⁵ For the closest choice, see naltrexone vs acamprosate.

Check for opioids and liver disease before the first dose

• Opioids. Naltrexone blocks opioid painkillers and can set off sudden, severe withdrawal in someone with opioids in their system. It is ruled out for people taking opioid pain medicine or on methadone or buprenorphine. Stop short-acting opioids, including tramadol, for at least 7 to 10 days before the first dose, and up to 2 weeks after methadone or buprenorphine.² Opioid cough and diarrhea medicines count too.² If you are planning surgery, see stopping naltrexone before surgery. After you stop naltrexone, an opioid dose you once took safely can be fatal: read the overdose warning.²
• Liver. The label warns of hepatitis and liver injury, and the APA advises against naltrexone in acute hepatitis or liver failure.^2,6 NICE suggests liver blood tests before starting.¹⁰ In a 2026 trial of 100 people with stable (compensated) alcohol-related cirrhosis, 64% on naltrexone were abstinent at 12 weeks against 22% on placebo, and no worsening of liver disease was attributed to the drug. It was one 12-week trial.²¹ Liver tests, what the results mean and the other cirrhosis studies are on naltrexone liver tests.
• Withdrawal. Naltrexone does not ease alcohol withdrawal.³ About half of people with AUD get some withdrawal symptoms when they stop drinking, and a small share need medical detox. Detox on its own is not treatment for AUD.⁵
• Mood. People taking naltrexone for alcohol should be watched for depression or suicidal thinking.² If you have thoughts of suicide, call or text 988.²²
• Pregnancy. The APA advises against AUD medicines during pregnancy or breastfeeding, except to treat withdrawal or another condition that needs medication.⁶

Early side effects, mostly nausea and headache, typically settle within 1 to 2 weeks, and the VA/DoD guideline suggests taking each dose with food.¹³ The opioid check, starting dose, timing and when to call are in starting naltrexone: the first two weeks.

Judge it at 4 to 6 weeks; courses run for months

Dose. Most people take one 50 mg tablet a day, the dose used in the trials behind its approval.² NICE starts at 25 mg a day and moves to 50 mg.¹⁰ Schedules and missed doses are covered on naltrexone dosage.

The first weeks. The VA/DoD calls naltrexone's side effects, "including initial transient nausea," generally minimal.¹³ See nausea and how long it lasts.

When to judge it. Neither the US label nor the guidelines say how many days it takes to start working. NICE gives a practical checkpoint: stop the drug "if drinking persists 4 to 6 weeks after starting."¹⁰

How long to stay on it. The tablet trials lasted 12 weeks, and other durations were not tested.² NICE says it should usually be prescribed for up to 6 months, "or longer for those benefiting from the drug who want to continue with it."¹⁰ NIAAA describes AUD treatment as lasting months, with medicines "generally not needed indefinitely."⁵ See how long does naltrexone take to work?

Medicine plus support beats medicine alone

The tablet label says naltrexone has not been shown to help except as part of a wider treatment plan.² That support can be light. In COMBINE, a 16-week US trial of 1,383 people, those given naltrexone plus brief check-ins with a clinician ("medical management") had more alcohol-free days than those given a placebo with the same check-ins: 80.6% vs 75.1%. Adding an intensive behavioral therapy did not beat naltrexone with check-ins.²³ COMBINE used 100 mg a day, twice the usual dose.^2,23

A 2025 Cochrane review found that adding a medicine to counseling probably cuts the share of heavy drinkers by about 10 percentage points, compared with counseling alone. Naltrexone was the most studied medicine in it.²⁴

No test predicts who it will help

Researchers have looked for traits that predict a good response. A 2014 review found that a family history of alcohol problems, and one variant of the mu-opioid receptor gene (OPRM1 Asn40Asp), went with better results in four of five and three of five studies. It also rated the studies at high risk of bias, and judged the data "still insufficient" to guide treatment.⁸ A 2020 meta-analysis of seven trials found no clear effect of that gene variant once it allowed for multiple comparisons.⁹

In the main Vivitrol trial, men and people who had stopped drinking before the first shot did better.¹⁵ For now, the way to find out is to take it and check progress with your prescriber at the point described in when to judge it.

Fewer than 3 in 100 people with AUD get medication

In 2024, 2.5% of people in the US aged 12 and older with alcohol use disorder (about 697,000 of an estimated 28.0 million) received medication for it.^11,25

NIAAA tells clinicians they need no special training or license to prescribe these medicines, and that they can be prescribed in primary care. It also notes that some patients hesitate because of misinformation they have heard about these medicines.^5,12 NIAAA calls all three FDA-approved medicines nonaddictive.¹ To find a prescriber, compare online programs that prescribe naltrexone or ask your primary care clinician.

More on naltrexone and alcohol

Share your experience

Have you taken naltrexone for drinking? Tell us what the first weeks were like, or read other people's naltrexone reviews.

How we checked

• Clinical claims come from the current tablet and Vivitrol labels (May 2026), the APA, VA/DoD and NICE guidelines, the 2023 AHRQ review, Cochrane reviews, trial abstracts and NIAAA pages.^{2-4,6,7,10,13,18,24,25}
• Each number was checked against our saved copy on October 17, 2026. The AHRQ and APA texts were read from cached copies; the Cochrane reviews and gene studies as abstracts.

Programs that treat alcohol use disorder

Online programs that prescribe naltrexone for alcohol, with their posted prices and what they include, are compared on where to get naltrexone online. Two app-based programs that ship compounded naltrexone, Reframe and Sunnyside, say on their own sites that their medication services are not meant to treat AUD.^27,28