89 Drugs Went Up Against Cocaine Addiction. None of Them Won.

If you’re on methadone or buprenorphine for opioid use disorder, there’s a medication with your name on it, decades of evidence behind it, and a real chance it keeps you alive. If you’re trying to stop drinking, naltrexone and acamprosate give you something to hold onto. If cocaine is the drug driving your life off the road, medicine has nothing to offer you. Not “nothing that’s been tried” — nothing that’s worked, out of everything that’s been tried.

Addiction medicine has spent thirty years and 89 different drugs trying to find a pill for cocaine use disorder, and the newest, largest look at that entire body of evidence says none of them work better than a sugar pill.

That’s the finding from a network meta-analysis published in the journal Addiction, led by the Bristol Evidence Synthesis Group — a methodology that lets researchers indirectly compare drugs that were never tested head-to-head by pooling every trial that compared each one to placebo. The team pulled together 163 randomized controlled trials, spanning 218 published reports and 14,871 participants, testing 89 different medications across eight drug classes: stimulant-replacement agents, antidepressants, anticonvulsants, dopamine agonists, and several others reached for over the years on the theory that cocaine’s dopamine-flooding mechanism should, in principle, be blockable. Across 12 separate effectiveness outcomes the researchers measured — things like sustained abstinence, reduction in days used, and retention in treatment — not one medication beat placebo. Not the best-performing drug in the pool. Not the combination therapies. None of them.

Why cocaine has resisted the model that worked for opioids and alcohol

The plain-language version of why this keeps happening: cocaine works on the brain through dopamine, the same chemical messenger involved in normal motivation, reward, and movement — which means almost any drug powerful enough to blunt cocaine’s effect also blunts things a person needs to function. Opioid use disorder medications work by occupying the same receptors heroin and fentanyl bind to, either fully (methadone) or partially (buprenorphine), which stabilizes a person without producing the same crash-and-craving cycle. Alcohol use disorder medications work through a few different, well-mapped pathways — naltrexone blocking the reward signal, acamprosate calming an overexcited nervous system in early sobriety. Cocaine doesn’t have an equivalent clean target. Every dopamine-pathway drug tested so far has run into the same wall: dial it down enough to blunt the high, and you’ve usually dialed down a person’s baseline mood and motivation right along with it, which is part of why so many of these trials show high dropout regardless of the drug’s theoretical mechanism.

This isn’t a story about doctors not trying hard enough. The trial count alone — 89 medications — represents a genuine, decades-long, good-faith search. A trade summary of the findings put it bluntly: the “drug-for-drug” approach that has produced real progress in opioid and alcohol treatment has, for cocaine, come up essentially empty. What’s left, and what the researchers point back toward, is what’s always been left: structured behavioral treatment — contingency management in particular, which pays people small incentives for verified abstinence and has a real evidence base behind it — alongside treatment for the co-occurring depression, ADHD, or trauma that often sits underneath stimulant use in the first place.

The contamination problem makes the treatment gap more dangerous, not less

None of this would carry the same urgency if cocaine still behaved the way it did in the 1990s. It doesn’t. CDC data show that roughly 80% of the 29,000 cocaine-involved overdose deaths in 2023 also involved fentanyl — meaning the person using cocaine, who has no reliable medication to help them stop, is now doing so inside a supply that’s frequently been cut with a synthetic opioid they never sought out and have no tolerance for. Nationally, overdose deaths are still trending down — provisional CDC data puts the count at roughly 69,147 for the year ending January 2026, a 13.2% decline — but that decline has been driven overwhelmingly by falling opioid deaths. Cocaine’s overlap with fentanyl means the two crises haven’t actually separated; a person with cocaine use disorder and no working medication option is, more than at almost any point in the last two decades, one contaminated bag away from an opioid overdose they didn’t choose.

None of this would carry the same urgency if cocaine still behaved the way it did in the 1990s.

For a provider or case manager sitting across from a client whose primary substance is cocaine or crack, the practical takeaway from this research isn’t “wait for a better drug.” It’s that the conversation should shift now, explicitly, toward contingency management programs where they’re available, integrated mental health treatment for co-occurring conditions, and — given the contamination data — the same fentanyl-supply-awareness conversation you’d have with an opioid client, because the line between the two populations is thinner than intake paperwork usually assumes. Eighty-nine medications is a lot of failed hypotheses. It’s also 89 reasons the next real breakthrough for cocaine use disorder almost certainly isn’t going to look like methadone. It’s going to have to look like something else entirely.