Ibogaine for Alcoholism

Alcohol was the third indication Howard Lotsof patented, in 1989, and it has been the least studied ever since. Almost everything claimed for it is borrowed from the opioid literature.

The one study

For decades the honest answer to this question was that nothing existed. That changed recently, and the change is small.

A pilot study of nine patients with moderate to severe alcohol use disorder was published in July 2026 in a Brazilian psychiatry journal. Patients were hospitalised with continuous monitoring and given escalating doses up to 400 mg. Five of the nine completed the protocol.

It matches the registered trial NCT03380728 in size, indication, institution and principal investigator: an escalating-dose trial at the University of São Paulo, completed in October 2024.

Not in design, though, and that is the most useful thing on this page.

It was registered as a randomised placebo-controlled trial

The registry keeps every version of an entry, so this is checkable by anyone.

The original entry spells out the intention in its own words: the first three patients would receive open-label doses, and if the three higher doses were tolerated, the next nine volunteers would receive those doses or placebo, randomly.

That placebo arm was never run. The trial that finished has one arm, no randomisation and no blinding.

What it reports on safety matters more than what it reports on drinking. Five of the nine had QT interval changes. Two required medication for psychomotor agitation, insomnia and raised blood pressure. Those are events in a hospital with continuous monitoring, which is not where most people take ibogaine.

An escalating-dose design also answers a different question from the one readers usually have. It asks how much can be given safely. It does not establish that any amount treats anything, and the study does not claim it does.

What the animal literature actually contains

The claim that ibogaine reduces drinking in animals is usually stated as though the evidence were broad. It is not.

The definitive systematic review and meta-analysis of ibogaine in animal models of addiction covers 27 studies. One of them is a behavioural study of alcohol.

That single study is genuinely interesting, and its most useful result is a negative one. When ibogaine, its metabolite noribogaine and the synthetic analogue 18-MC were compared directly, noribogaine raised GDNF and reduced alcohol self-administration when infused into the relevant brain region, as ibogaine does. 18-MC did neither.

That matters here because 18-MC is the compound designed to keep the anti-addictive effect while dropping the cardiac risk. On the one alcohol experiment that compared them, it did not keep the effect. 18-MC sets out what follows from that.

Where the numbers people quote come from

If you have seen a percentage attached to ibogaine and alcohol, it almost certainly traces back to a patent rather than a study.

US 4,857,523, “Rapid method for attenuating the alcohol dependency syndrome”, was granted in 1989. Like its predecessors it asserts effectiveness without a trial behind it, and the first patent in the series lists a single reference: an entry in a chemical index.

What ibogaine is sets out how that sequence of patents came to stand in for evidence, and why there is no success rate covers what a real outcome figure would have to look like.

Enlarge

The comparison a reader is really making

Someone reading this page is usually weighing ibogaine against either doing nothing or against conventional treatment. Both comparisons go badly for ibogaine, and for different reasons.

Against conventional treatment, the asymmetry is stark. Alcohol dependence has medications with randomised trial evidence behind them, and psychosocial treatments with the same. Ibogaine has nine people, five of whom finished.

Heavy drinking also damages the liver, and ibogaine is cleared by a liver enzyme whose activity already varies several-fold between people. Noribogaine explains why that variation matters and why the drug is still present days after the experience ends.

The honest summary

There is no evidence that ibogaine treats alcohol dependence, and the absence is not a technicality. No randomised trial has been run. The one study that set out to treat alcohol dependence enrolled nine people, was not designed to measure efficacy, and lost four of them. The animal case rests on a single behavioural experiment. The percentages in circulation come from a patent application.

What the field does have is a good safety observation from a clean study: even under continuous hospital monitoring, in nine people, five had QT changes. Ibogaine and the heart is where that leads.