Cocaine Killed More People Than Ever Last Year. There Still Isn’t a Pill for It.

On June 19, 1986, Len Bias, a 22-year-old forward the Boston Celtics had drafted two days earlier, used cocaine at a party in his college dorm and went into cardiac arrest. He was dead within hours. Congress had a bill on the floor within weeks. By the fall, the Anti-Drug Abuse Act of 1986 had created the 100-to-1 sentencing disparity between crack and powder cocaine — a ratio a federal commission would later find had no basis in pharmacology, and that a generation of mostly Black defendants would serve mostly White defendants’ sentences under. That law took 24 years to even partially fix.

We know how to pass a law fast when a famous person dies of cocaine. We have never once managed to fund a medicine for it.

Forty years later, cocaine is having its worst year on record, by the only measure that matters: bodies. Cocaine-involved overdose deaths have more than quadrupled since 2015, and cocaine was present in roughly 28% of all U.S. overdose deaths in the most recent full year of data — a number that has climbed even as the overall overdose count has fallen for two straight years. The country is, by every headline, winning the opioid fight for the first time since the crisis began. Nobody is winning the cocaine fight, because almost nobody is fighting it. There is still no FDA-approved medication for cocaine use disorder — not one, not ever — while the fentanyl era has quietly turned cocaine into a delivery mechanism for a drug most of the people who die from it never chose.

If you use cocaine now, in 2026, you are more likely than at any point in this crisis to also be using fentanyl — usually without deciding to. That’s not a scare line. It’s what CDC surveillance data shows: fentanyl contamination now runs through cocaine, methamphetamine, and counterfeit pills alike, and nearly half of all overdose deaths now involve both an opioid and a stimulant together. Cocaine didn’t get more dangerous because of what it is. It got more dangerous because of what else is in the bag.

The supply shock nobody priced in

The economics are almost boringly simple, which is what makes them worth explaining plainly. Colombian coca cultivation cratered between 2000 and 2013 — from roughly 168,000 hectares down to about 48,000 — under an aggressive, U.S.-backed eradication campaign. Then, starting around 2015, aerial fumigation stopped, a peace deal with FARC guerrillas emptied out the armed groups that had been (imperfectly) suppressing cultivation in some regions, and a crop-substitution program meant to move farmers off coca largely backfired. By 2022, cocaine output was more than three times its 2015 level. Economists Xinming Du, Benjamin Hansen, Shan Zhang, and Eric Zou modeled what that supply shock did on the demand side in the U.S. and estimated roughly 1,500 additional annual overdose deaths attributable to the post-2015 surge alone. University of Oregon economist Ben Hansen, one of the paper’s authors, put the mechanism in plain terms: “Because cocaine is an experience good, if you have a big supply shock, that leads to more people potentially using it.” More coca, more cocaine, lower prices, more first-time and returning users. That part of the story is a hundred years old — supply shapes demand, not just the other way around.

What’s new is the part that isn’t about coca at all. The U.S. synthetic opioid supply matured in parallel with the cocaine glut, and the two markets collided. Distributors moving product through the same networks, the same cutting tables, the same low-level street sellers, don’t always keep their inventories clean — sometimes deliberately, more often through simple cross-contamination or an attempt to stretch supply with whatever’s cheap and potent. Fentanyl is both. The result, according to the European Union Drugs Agency’s 2026 cocaine report, is that drug-checking services on two continents are now finding cocaine samples at purities above 80% — a number that sounds like good news for a buyer and is actually a warning sign, because high-purity batches are frequently the ones later found to carry fentanyl or other adulterants riding along for the trafficker’s convenience, not the user’s.

synthetic opioid supply matured in parallel with the cocaine glut, and the two markets collided.

What Len Bias’s law actually built, and what it didn’t

Here is the through-line worth sitting with: in 1986, Congress responded to a single, telegenic cocaine death by building an entire sentencing architecture around the form cocaine took — crack, associated in the press and in Congress with Black urban communities, got mandatory minimums for five grams; powder, associated with white suburban and college use, needed 500 grams to trigger the same sentence. The U.S. Sentencing Commission itself would later conclude the 100-to-1 ratio “significantly undermines” the credibility of federal drug policy. The Fair Sentencing Act of 2010 cut that ratio to 18-to-1 — still not 1-to-1, still not evidence-based, just less indefensible. That fight took 24 years and thousands of federal sentences served under a ratio Congress’s own experts said made no pharmacological sense.

Now do the comparison that should embarrass everyone who works in this field: it took Congress six weeks to write a criminal law in response to one death. It has taken the NIH and FDA more than three decades to fail to produce a single approved medication for cocaine use disorder — not because the biology is unreachable, but because cocaine use disorder has never had a Len Bias moment aimed at treatment instead of punishment. We regulate cocaine like it’s 1986. We treat it like it’s still nobody’s problem to solve.

The one thing that works is the thing nobody will pay for

That’s not quite true — there is one treatment with real evidence behind it. Contingency management, which pays people small, escalating incentives for biochemically verified abstinence, is the only behavioral intervention with strong outcome data for stimulant use disorder, cocaine included. A national cohort study of veterans published in the American Journal of Psychiatry found that patients with stimulant use disorder who received contingency management were 41% less likely to die in the year after starting treatment than matched patients who didn’t. Not less likely to relapse — less likely to die. That is as strong a mortality signal as anything in this field has produced in years, for a population medicine has otherwise shrugged at.

If you have ever sat across from a case manager and been told “there’s really nothing we can prescribe you for this one,” you already know what that gap feels like from the inside — like being handed a map with your destination erased. Contingency management is one of the only things clinicians can offer instead of a shrug, and it still isn’t standard anywhere. Only a handful of states have moved to cover it under Medicaid, and most private insurers still don’t reimburse it at all, partly because federal anti-kickback rules were written before “paying someone a gift card to stay alive” was a clinical intervention anyone anticipated regulating, and untangling that took years state programs didn’t have. Arizona is one of the states that has not applied for Medicaid coverage of contingency management. That is not a footnote. That is the treatment gap, in one sentence, sitting in our own backyard.

The number that should be driving funding isn’t

The cruelest part of this story is timing. The country is, right now, in the middle of the first sustained multi-year decline in overdose deaths since the crisis began — 69,147 deaths in the twelve months ending January 2026, down 13.2% from the year before, largely on the strength of falling fentanyl potency and years of naloxone saturation finally paying off. That decline is real, it is hard-won, and it deserves the credit it’s getting. But a crisis narrative built entirely around fentanyl potency has a blind spot exactly the size of cocaine’s rising share of the death toll. Fentanyl getting less deadly on its own doesn’t help the person whose cocaine has fentanyl mixed into it without their knowledge. It doesn’t touch the population using stimulants who were never opioid-tolerant to begin with, for whom a small, unlabeled dose can be the first and last exposure of their life.

But a crisis narrative built entirely around fentanyl potency has a blind spot exactly the size of cocaine’s rising share of the death toll.

Cocaine use disorder is not a niche problem waiting politely at the back of the line. It is 28% of a death toll that is otherwise, correctly, being reported as good news — and it is the substance class where American medicine has made the least progress in forty years, not because the science is stuck, but because nobody has funded it like it matters. Contingency management works. Almost nobody pays for it. That is a policy choice, made new every budget cycle, not a scientific limitation.

We have watched this exact failure of imagination before — a fast, punitive response to a cocaine death, and a slow, underfunded response to cocaine treatment. The names change. Len Bias in 1986. Whoever the CDC will count, unnamed, in this year’s provisional data. The country still hasn’t decided that the second kind of death deserves the first kind of urgency. Until it does, the medicine cabinet for the second most common overdose substance in America stays empty, and the treatment that actually works keeps waiting on a Medicaid application nobody in Arizona has filed yet. Find contingency management and stimulant treatment programs near you →