Naltrexone vs Acamprosate: Which Works Better?

Side-by-side comparison

Which drug fits you?

The American Psychiatric Association (APA) recommends offering either drug to people with moderate to severe alcohol use disorder, whether the goal is to drink less or to stop.⁵ Your health and circumstances usually decide between them. This table pulls together who shouldn't take which drug, and what the labels, the APA, the VA/DoD guideline and a SAMHSA advisory say about specific situations. Use it to prepare for a talk with a prescriber.

The VA/DoD points in this table come from the guideline's discussion text, not its graded recommendations. In those recommendations, naltrexone sits in the higher-rated tier and acamprosate in the lower one; see What the guidelines say.⁶

Which one works better?

There is no clean winner, partly because direct comparisons are scarce.

When a 2014 JAMA meta-analysis pooled the trials that did compare them directly, it found "no statistically significant difference between them for return to any drinking (RD, 0.02; 95% CI, -0.03 to 0.08) or heavy drinking (RD, 0.01; 95% CI, -0.05 to 0.06)."¹ The 2023 update of that review, done for the federal Agency for Healthcare Research and Quality (AHRQ), reported that "no new eligible studies provided head-to-head comparisons."²

Against placebo, both work, modestly. The 2023 review expressed results as a number needed to treat: how many people need the drug for one extra person to benefit.

Source: McPheeters et al, JAMA 2023.⁹ The wide confidence intervals mean the true numbers could be quite different. The authors concluded that the evidence supports "oral naltrexone at 50 mg/d and acamprosate as first-line pharmacotherapies for alcohol use disorder."⁹

The Cochrane reviews, both from 2010, split along similar lines. Across 50 trials and 7,793 patients, naltrexone cut the risk of heavy drinking to 83% of the placebo group's risk, while its effect on any return to drinking "missed statistical significance."¹³ Across 24 trials and 6,915 people, acamprosate significantly reduced the risk of any drinking (risk ratio 0.86) and lengthened abstinence, but its effect on heavy drinking did not reach significance.¹⁴

That pattern explains the common shorthand: acamprosate for staying sober, naltrexone for cutting down. A 2013 meta-analysis of 64 placebo-controlled trials backs it, but only as a tilt. It found acamprosate "slightly more efficacious in promoting abstinence and naltrexone slightly more efficacious in reducing heavy drinking and craving."¹² The APA recommendation covers both drugs for both goals.⁵

The naltrexone label itself calls the effect modest. In one of its placebo-controlled studies, with 104 patients, abstention rates were 51% on naltrexone against 23% on placebo, but the label says "the expected effect of the drug is a modest improvement in the outcome of conventional treatment."³

The COMBINE trial

COMBINE is the trial people cite when they say acamprosate doesn't work. It enrolled 1,383 volunteers who had recently stopped drinking, at 11 US academic sites. They got 16 weeks of naltrexone (100 mg a day), acamprosate (3 g a day), both, or placebo, with or without a combined behavioral intervention (CBI) delivered by specialists.⁸

Naltrexone reduced the risk of a heavy drinking day (hazard ratio 0.72). "Acamprosate showed no significant effect on drinking vs placebo, either by itself or with any combination of naltrexone, CBI, or both."⁸

One large trial is not the whole record. The pooled reviews above cover dozens of trials, and they found acamprosate helped prevent a return to any drinking.^9,14 COMBINE also used naltrexone at 100 mg a day, twice the usual label dose.^3,8 Our reading: COMBINE shows acamprosate added nothing measurable for that trial's US volunteers, under its conditions. It doesn't cancel out the pooled evidence.

What the guidelines say

The APA says either drug should be offered to people with moderate to severe alcohol use disorder who want to cut down or stop. The other conditions: they prefer medication or haven't responded to other treatment, and they have no contraindications. The APA's reason: "Naltrexone and acamprosate have the best available research evidence as pharmacotherapy for patients with AUD."^5,16

England's NICE guideline says to consider either drug after withdrawal for moderate to severe dependence, alongside psychological treatment. For harmful drinking and mild dependence, it rates the evidence for acamprosate as weaker than the evidence for naltrexone.¹⁵

The VA/DoD puts the two drugs in different tiers: naltrexone in its top tier with topiramate, acamprosate in a lower tier with disulfiram. Its "insufficient evidence to recommend for or against the routine use of one of these medications over another" applies within a tier, such as naltrexone versus topiramate, not to naltrexone versus acamprosate.⁶ Topiramate, disulfiram and the other medicines for alcohol use disorder are compared on their own page.

Do you have to stop drinking first?

For acamprosate, yes. The label limits it to people "who are abstinent at treatment initiation," and adds that its efficacy "has not been demonstrated in subjects who have not undergone detoxification and not achieved alcohol abstinence."⁴ Three of its approval studies enrolled 998 patients who had been through inpatient detox and were sober on the day they were randomized. A fourth study, in people who hadn't detoxed, "failed to demonstrate superiority of acamprosate calcium over placebo."⁴ The label also says treatment should start "as soon as possible after the period of alcohol withdrawal" and continue if the person relapses.⁴

The naltrexone tablet label sets no abstinence condition in its indication.³ Vivitrol, the injectable form, does: patients "should not be actively drinking at the time of initial VIVITROL administration."⁷

Neither drug eases withdrawal. Acamprosate "does not eliminate or diminish withdrawal symptoms," and the Vivitrol label says the same of naltrexone.^4,7 People who are physically dependent on alcohol need a plan for withdrawal before either medicine.

Mood: both labels ask for monitoring

In acamprosate trials, suicide-related events were more common on the drug than on placebo: 1.4% vs 0.5% in studies of six months or less, and 2.4% vs 0.8% in year-long studies.⁴ The naltrexone label says depression and suicidal thoughts have been reported after marketing in people treated for opioid dependence, with no causal link shown.³ Both labels say patients should be monitored for depression or suicidal thinking, and that families and caregivers should watch for it too.^3,4

Side effects

Diarrhea is acamprosate's signature side effect. In the label's pooled trials it affected 16% of people on the drug and 10% on placebo.⁴ The 2010 Cochrane review found it was the only side effect reported more often on acamprosate than placebo.¹⁴

For naltrexone, the problems tend to be in the stomach and the head. The 2023 review found nausea (risk ratio 1.73) and vomiting (1.53) more common than on placebo, along with dizziness and insomnia; the Cochrane review also flagged daytime sleepiness.^2,9,13 In the few head-to-head studies, people on acamprosate had "a slightly lower risk of headache and vomiting" than people on naltrexone.² For how often each one occurs, see naltrexone side effects.

Pill count and cost

Six tablets a day is a burden, and the 2023 review counted it against acamprosate, noting that it and topiramate are "less convenient to take, with multiple doses per day."² Naltrexone is one tablet, or one shot a month.^3,7

Medicaid coverage favors naltrexone too: in 2023, 49 of 53 fee-for-service programs covered the tablets and 36 covered acamprosate.¹⁰ What the online programs charge is on our naltrexone cost page.

Can you take them together?

Yes, from a safety standpoint. The acamprosate label says "clinically important interactions between naltrexone and acamprosate were not observed."⁴

Whether it helps is another matter. In COMBINE, "no combination produced better efficacy than naltrexone or CBI alone in the presence of medical management."⁸

Adding a medicine to counseling looks more useful than adding a second medicine. A 2025 Cochrane review found that medication plus psychosocial treatment "probably" reduces the number of heavy drinkers compared with psychosocial treatment alone (an absolute difference of 10 percentage points, moderate-certainty evidence). Against medication alone, adding counseling "may have little to no effect" on abstinent days, and its effect on heavy drinking was uncertain (low- to very-low-certainty evidence). Naltrexone was the most studied drug in that review.¹⁷

Most people get neither

Whichever drug is the better fit, most people with alcohol use disorder get no medication at all. 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.¹⁸ If you decide to try naltrexone, you can get a prescription from your own clinician or an online program.