The Sinclair Method: What It Is and What the Evidence Shows

The protocol

Sinclair laid out the rules in a 2001 review in Alcohol and Alcoholism: give naltrexone to people who are still drinking, tell them to take it only when they expect to drink, and continue the treatment indefinitely.¹ The Finnish clinics he describes put it more bluntly: "take naltrexone 1 h before drinking: no drinking without taking the naltrexone first, and no naltrexone if no alcohol is drunk that day."¹

That is the whole protocol: no detox first and no requirement to quit. In the US it is off-label, because the tablet label's dose is 50 mg once a day.⁶

Sinclair treated drinking as a learned response that alcohol keeps reinforcing. Naltrexone, an opioid blocker, is meant to blunt that reinforcement, so each drink taken on the drug should weaken the habit a little. He called this extinction and argued that it weakens only the responses made while reinforcement is blocked, which is why the pill goes with the drink. He pointed to his group's animal studies, in which opioid blockers worked when paired with drinking but not when given during abstinence.¹

Extinction is Sinclair's explanation; it is not an established mechanism for alcohol. The FDA label for naltrexone tablets says the drug's mechanism in alcoholism is not understood, though preclinical data point to the body's own opioid system.⁶

What the first months are like

Sinclair expected change to come slowly. In the open-label Finnish program he describes, which had no control group, reported craving fell steadily over the first 110 days of treatment, and he described craving and drinking in the trials he cites as falling slowly over many weeks.¹ He also wrote that in controlled trials the gains over placebo seen in the first 3 months held with longer treatment but did not grow much further.¹

The controlled trials of targeted use ran for 8 to 12 weeks, and one for 32, so they say little about how the method plays out over years.^2,4 Drinking continues during this time by design; what that means for intoxication is under safety.

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.⁹ The opioid check, starting dose, timing and when to call are in starting naltrexone: the first two weeks.

What controlled trials found

Randomized, placebo-controlled trials of targeted naltrexone exist, but they don't test the Sinclair Method exactly. Each defined targeting a little differently, and most ran for 8 to 12 weeks.

Planning itself may help. In the 2003 trial and its reanalysis, simply having a pill to take before risky situations helped, even when the pill was a dummy.^3,7 Part of what people credit to the method may be the act of deciding in advance.

The effects are modest. The 2009 authors concluded that their results called for additional strategies or more effective medicines than naltrexone.⁴

Counseling mattered. In the Finnish trial, which Sinclair co-authored, naltrexone beat placebo only in the groups that also got coping-skills therapy.² Santos and colleagues also paired the drug with weekly counseling.⁵

One longer Finnish study gave naltrexone, acamprosate or disulfiram for 12 weeks under supervision, then switched to targeted use for up to 52 weeks. It was open-label, so patients knew their drug. During the targeted phase, naltrexone and acamprosate did not differ on drinking outcomes, and disulfiram patients had more abstinent days.¹¹

The wider evidence on naltrexone

The big reviews of naltrexone mostly reflect daily dosing, and they show a real but limited benefit. A Cochrane review found naltrexone cut the risk of heavy drinking to 83% of the risk on placebo.¹² The 2023 federal review in JAMA concluded that its findings "support the use of oral naltrexone at 50 mg/d and acamprosate as first-line pharmacotherapies for alcohol use disorder."¹⁰ That conclusion is for 50 mg taken every day. Acamprosate and the other medications for alcohol use disorder are compared on a separate page.

The review closest to the Sinclair Method's target group is less encouraging. In 2018 Palpacuer and colleagues pooled 32 double-blind trials, with 6,036 patients, in people who were still drinking. On total alcohol consumption, "No efficacy was observed for naltrexone or acamprosate."¹³ That analysis mixed daily and targeted trials, so it doesn't settle the question, but it does undercut claims of a dramatic effect. We found no meta-analysis devoted to targeted naltrexone.

Is it FDA-approved?

Naltrexone tablets are FDA-approved for alcohol dependence. The Sinclair Method's way of taking them is not. The label says "A dose of 50 mg once daily is recommended for most patients."⁶ Taking it only on drinking days is therefore off-label in the US, meaning it departs from the dosing the label describes. Two US online programs describe as-needed use. Oar Health sells Clutch, a compounded mint it markets for targeted naltrexone use, and Ria Health's FAQ says people on naltrexone typically take it an hour before they expect to drink.^14,15 Both are off-label uses, and Clutch, as a compounded drug, is not FDA-approved.¹⁴

The as-needed idea does have an approved cousin in the UK, using a different drug. The UK's National Institute for Health and Care Excellence (NICE) recommends nalmefene, another opioid blocker, for reducing drinking in people with alcohol dependence and a high drinking-risk level, and only alongside continuous psychosocial support.¹⁶ That guidance is about nalmefene and does not endorse the Sinclair Method. A 2016 meta-analysis put nalmefene's effect on heavy-drinking days at a Hedges' g of −0.20, a small effect.¹⁷

None of the US guidance we read mentions targeted dosing. The American Psychiatric Association recommends offering naltrexone to patients with moderate to severe alcohol use disorder, and the VA/DoD guideline gives it a "Strong for" rating, but neither summary describes an as-needed schedule.^8,9

Safety: what the method doesn't change

Taking naltrexone less often doesn't switch off its warnings. Common side effects have their own page; the three points below bear on this way of taking it.

You can still get drunk. "Opioid antagonists do not block the motor impairment from ethanol, and in some conditions can even increase the intoxication," Sinclair wrote, and he said patients should be warned strongly about the dangers of intoxication, for example when driving.¹ The label adds that naltrexone is not an aversive therapy and does not cause a disulfiram-like reaction when you drink.⁶

Opioids are the big risk. Naltrexone is contraindicated with opioid pain medicine and in people who are dependent on opioids. The label warns that trying to overcome the blockade by taking opioids "may lead to life-threatening opioid intoxication or fatal overdose," and that after stopping, people may be more sensitive to opioids, even at lower doses.⁶ Our page on how long naltrexone stays in your system covers timing.

Don't double up. The label's dose is 50 mg once a day, and it warns of "a higher risk of hepatocellular injury with single doses above 50 mg."⁶ The method uses one tablet before a drinking session; a second tablet the same day is covered in the FAQ.

Where the 78% comes from

"Of the 147 patients who had completed enough treatment to be classified, 115 (78%) were considered successful and had a mean (± SEM) final drinking level of 9.4 ± 1.0 drinks weekly; 38 of the patients (26%) reached abstinence, although only 3% listed it as the goal of treatment."¹

The sentence is from Sinclair's 2001 review. It describes an open-label trial at two Finnish clinics that gave targeted naltrexone without detox first, drawn from an ongoing, expanding program.¹ It leaves out four things a success rate needs:

A rate counted only among people who stayed leaves out everyone who dropped out, including anyone who quit because it wasn't working. None of that makes the program a failure. It means 78% can't be read as the chance that the method will work for you.

The figure is still in use. Contral Clinics, a Helsinki-based treatment provider whose website cites Sinclair's 2001 paper, says "78% of our clients succeed in their goal of reducing their alcohol consumption."¹⁸ At least one provider has stepped back from it. Rethink Drink, a UK service that prescribes the method, now writes: "We used to quote the 78% ourselves, like most people in this field."¹⁹

The same paper reports a long-term follow-up, about 31 months after patients started. Of the 27 responders who said they were still taking naltrexone when drinking, all said they drank less than before treatment; only 8 of the 30 who had stopped following the instructions said the same.¹ That is self-report from people who chose to answer, with no comparison group.

Counseling and the abstinence question

Sinclair was specific about the kind of counseling. He wrote that "naltrexone should not be used together with supportive therapy enforcing abstinence," and that clinics should use cognitive behavioral therapy (CBT) with it, ideally structured CBT programs shown to work with naltrexone.¹

His reasoning came from a pattern across three trials: naltrexone helped alongside coping-skills therapy, which accepts that slips may happen, but not alongside therapy that pushed for complete abstinence.¹ The Finnish trial found the same split.² The pattern held across those trials, but it comes from subgroup comparisons, and no guideline we read has turned it into a rule.

A note on the method's origins

Sinclair worked at Finland's National Public Health Institute when he wrote the 2001 review.¹ Its reference list includes his own 1989 US patent, Method for Treating Alcohol-Drinking Response (US Patent 4,882,335).¹ Holding a patent doesn't make the work wrong. It is the kind of interest readers should know about when the main source for a claim is the method's inventor.

He also noted a practical upside: taking the drug only on drinking days lowers the monthly cost of medication.¹ For current prices, see how much naltrexone costs.