Meth and Cocaine Now Kill More Americans Than Opioids. Medicine Still Has Nothing to Offer.

Jeffrey Steen was working on a doctorate in social work when he tried crystal meth for the first time, a few months after moving to New York City for the program. It worked, at first, better than almost anything else in his life had. “For a while at least, meth seemed to aid me in just about all aspects of my life,” he wrote later in Filter — it dissolved his social anxiety and filled in for an undiagnosed mental health condition nobody had named yet. Within a couple of years he’d gone from smoking it occasionally to injecting it daily. He ended up in an Upper East Side hotel with a plan to die there. He is alive because a man named Jake, met on Grindr and further along in his own recovery, found him and got him to a hospital.

In 2025, meth and cocaine combined killed more Americans than opioids for the first time on record — 16,470 deaths (8,951 methamphetamine, 7,519 cocaine) against 15,914 opioid deaths — per Millennium Health’s 8th Annual Signals Report. Researchers reading that data are calling it a fifth wave of the overdose crisis.

The overdose crisis didn’t shrink. It changed drugs — and nobody built a treatment for the ones it changed into.

That’s Rize’s position, and the data backs it without much room for hedging. Opioid deaths are genuinely, meaningfully down — more on that below, because the decline is real and it matters. But the absence of a stimulant medication means that drop isn’t translating into fewer overdose deaths overall — it’s translating into a different, less-treated crisis wearing the same statistical column. This isn’t the first time the country has watched a stimulant surge outrun its treatment system. It’s arguably the second. Last time, the response was prison, not medicine, and the country is still arguing about what that cost.

Steen recovered, and he finished the doctorate, with a combination nobody handed him in a program brochure: drug checking, syringe safety, and a weekly harm reduction therapy group, stacked underneath a twelve-step program. Nobody gave him a prescription that made the craving quiet down, because none exists. There is still no FDA-approved medication for methamphetamine or cocaine use disorder — not one, for either drug, despite decades of trials. Steen’s recovery ran entirely on behavior, community, and time. That’s not a footnote. It’s the whole architecture of stimulant treatment in America, about to be tested at a scale it was never built for.

What 1.69 Million Drug Tests Are Actually Telling Us

Millennium Health’s report draws on 1.69 million urine drug tests collected from 530,000 people in substance use disorder treatment, across all 50 states, from 2016 to 2025. Eighty-five percent of samples that tested positive for fentanyl also tested positive for a stimulant — the highest co-positivity rate the lab has recorded. “Clinicians are managing increasingly complex patterns of polysubstance use without adequate therapeutic tools,” Angela Huskey, PharmD, Millennium Health’s Senior Vice President and Chief Clinical Officer, said of the findings.

To see why “fifth wave” is the word researchers reached for, the wave metaphor is worth being precise about. The first wave was prescription opioids in the late 1990s. The second was heroin, roughly 2010 onward, as pill supplies tightened and people already dependent moved to a cheaper, more available drug. The third was fentanyl — the synthetic opioid 50 to 100 times more potent than morphine — flooding the heroin and counterfeit pill supply starting around 2013 and driving deaths to levels the first two waves never approached. The fourth wave — the one a peer-reviewed analysis in Substance Use & Misuse tracking data from 1999 to 2023 puts a hard number on — was fentanyl combining with stimulants: the share of fentanyl-involved deaths that also involved a stimulant rose from 7.95% in 1999 to 56.67% by 2023. That same study found methamphetamine deaths rose from 547 in 1999 to 34,855 in 2023, and cocaine deaths rose from 3,822 to 29,449 over the same stretch — and the burden isn’t distributed evenly: methamphetamine deaths are disproportionately concentrated among American Indian and Alaska Native people (44.5% of that group’s overdose deaths involved meth), and cocaine deaths are disproportionately concentrated among Black Americans (48.47%).

The second was heroin, roughly 2010 onward, as pill supplies tightened and people already dependent moved to a cheaper, more available drug.

So what makes 2025 different enough to earn a new number? Two things. First, the ratio flipped: stimulants aren’t just riding alongside fentanyl anymore, they’re now killing more people than opioids in absolute terms, which hasn’t happened before in this crisis. Second, the growth is accelerating specifically among people who aren’t using fentanyl at all — methamphetamine detections rose 4.5% nationally in 2025 among non-fentanyl users, and cocaine detections rose 13.5%, steepest in the South. That’s a population increasingly using stimulants as a primary drug, not a fentanyl side effect. A CDC data brief covering 2003 through 2023 shows the same divergence in the government’s own numbers: the overall overdose death rate fell 4.0% from 2022 to 2023, driven by declining synthetic opioid and heroin deaths — the first such opioid decline since the surge began around 2013 — while cocaine death rates kept climbing (up 4.9%) and methamphetamine and other psychostimulant death rates kept climbing too (up 1.9%). Two epidemiological lines, moving opposite directions, layered on top of each other.

Here’s the honest caveat: it is not settled that “fifth wave” describes something genuinely new versus the fourth wave deepening under a catchier name. The Substance Use & Misuse authors frame the fentanyl-stimulant convergence itself as the fourth wave and don’t use “fifth” at all; Medscape’s sourcing treats the flip in absolute death counts as the new threshold. Both readings can be true of the same data. What isn’t in dispute is the substance: stimulants are now the deadliest drug category in America in a year when opioid deaths fell, and the clinical toolkit for treating stimulant use disorder hasn’t changed. Whatever number you put on the wave, that gap is the story.

America Already Ran This Experiment, and Chose Prison

This is not the first time a stimulant surge arrived with no medical answer. Crack cocaine moved through American cities in the late 1980s with the same absence of pharmacotherapy meth and cocaine have now — no medication then, none for cocaine today. What the country built instead was the Anti-Drug Abuse Act of 1986: a 100-to-1 sentencing disparity, so five grams of crack drew the same five-year mandatory minimum as 500 grams of powder cocaine, chemically almost the same drug. Congress called crack uniquely addictive and violent, claims that didn’t hold up scientifically but held for two decades politically. By 2009, 79% of people convicted under the crack provisions were Black, though use was roughly proportional across races. The Fair Sentencing Act cut the ratio to 18-to-1 in 2010 — two decades after the damage was done.

That history isn’t a tangent. It’s the pattern repeating in slow motion. A stimulant crisis with no medical treatment gets met, historically, with punishment aimed at the people using and selling rather than treatment aimed at the disorder driving the use — and the punishment lands hardest on communities with the least power to resist it. Nobody in 2026 is proposing a new 100-to-1 law. But every dollar routed to enforcement instead of contingency management, every jurisdiction that treats a positive stimulant test as a probation violation instead of a treatment referral, is a smaller-scale version of the same choice: manage the supply, ignore the disorder. The question in front of policymakers now is whether this stimulant wave gets a treatment system or a courtroom. The 1980s already showed which one is faster to build and which one actually works.

Zero Medications, One Underused Answer

Here’s the plain fact underneath all of it: there is no pill for this. Not for methamphetamine use disorder, not for cocaine use disorder. Buprenorphine and methadone work for opioid use disorder because they act on the same brain receptors opioids do, easing withdrawal and blunting the reward of using again; naltrexone blocks those receptors outright. Stimulants work on a different chemical system — mainly dopamine, the brain’s motivation-and-reward messenger — and despite dozens of compounds tested in clinical trials over decades, nothing has cleared the bar for FDA approval. The FDA issued its first draft guidance in October 2023 just to help drugmakers design stimulant use disorder trials properly — a sign of how far behind this field is, not how close to a breakthrough.

Here’s the plain fact underneath all of it: there is no pill for this.

Into that vacuum sits one behavioral treatment with a genuinely strong evidence base: contingency management, which is really just this — paying people small, tangible rewards, like vouchers or prize drawings, for turning in a stimulant-negative urine test. It sounds almost too simple to be a medical intervention, and it draws real political resistance for exactly that reason; “the state can’t pay people not to use drugs” is a sentence that has killed funding proposals in more than one legislature. But the evidence holds up, and it’s currently the strongest tool the field has for stimulant use disorder, according to Acuity News’s review of the treatment landscape. The federal government has started, cautiously, to pay for it: the Centers for Medicare & Medicaid Services has approved Section 1115 Medicaid waivers for contingency management programs in five states — California, Delaware, Hawaii, Montana, and Washington — with Michigan and Rhode Island applications pending, per KFF’s tracking. California’s Medi-Cal Recovery Incentives Program, the largest and longest-running, has around 8,500 members enrolled across roughly 100 sites, and the state’s own reporting shows better treatment retention, fewer emergency department visits, and fewer overdose deaths among participants. That program also has a deadline: California’s Medicaid director must make the case for its value convincingly enough by the end of 2026 to keep the federal funding that sustains it, in a political climate not especially friendly to programs that look, on paper, like the government handing out gift cards.

You know what this gap feels like if you’ve sat in a program built around medication schedules — Suboxone check-ins, methadone dosing windows — while your drug of choice was crystal, not pills, and the best anyone offered was “behavioral support” and a pamphlet. That’s not a smaller problem than opioid treatment. It’s the same problem with none of the tools. If you’re the case manager or clinician reading this: pull your intake screening form this week and check whether “stimulant use” routes a client toward an actual contingency management referral or just generic counseling — and if your state isn’t one of the five with an approved waiver, put that question on the next team meeting’s agenda, because the funding mechanism exists somewhere for someone to go find.

The Supply Didn’t Just Grow. It Flooded.

None of this is happening only because of demand inside the U.S. It’s also a supply story, and a global one. Global cocaine production hit an estimated 4,000-plus tonnes of pure product in 2024 — roughly four times what it was a decade earlier — per the UN Office on Drugs and Crime, with quality rising and prices falling as organized crime expands into new and existing markets. Methamphetamine trafficking, tracked through global seizure data, is growing roughly 13% a year, with production spreading into Africa, the Middle East, and parts of Europe that weren’t previously major markets. “We have seen an unprecedented spike in new types of drugs on the market, and worryingly, some are more potent or dangerous than before,” UNODC Executive Director Monica Juma said when the report was released. More supply, cheaper and purer, arriving into a domestic system with no medication and a criminal-legal default from the last time this happened — that’s the mechanism, not a mystery.

It also explains something about the co-use numbers that’s easy to miss if you only look at national totals. Millennium Health’s regional breakdowns show fentanyl-methamphetamine co-positivity jumping 40.7% in the South in a single year, and fentanyl-cocaine co-positivity rising 19.3% in the Northeast; over the 2020-to-2025 window, samples showing cocaine and methamphetamine together alongside fentanyl rose 249% in the West and 239% in the Northeast. Regional drug markets are converging on the same chemistry at different speeds, which is exactly what you’d expect from a supply shock moving through existing distribution networks rather than a single national trend arriving everywhere at once.

It also explains something about the co-use numbers that’s easy to miss if you only look at national totals.

If you use stimulants, or you’re early in recovery from them, here’s one thing worth being precise about, because getting it wrong could cost someone their life: naloxone does not reverse a methamphetamine or cocaine overdose. It works only on opioids. What it does do — and this is not nothing — is reverse the opioid piece of a mixed overdose, and with 85% of fentanyl-positive samples nationally also testing stimulant-positive, the stimulant supply and the fentanyl supply are, practically speaking, the same supply in most of the country now. Carrying naloxone, knowing how to use it, and having someone nearby who knows you use are still real, free protections tonight — genuinely useful, and not a solution to the stimulant problem. Nobody should tell you it is.

Waves are a metaphor built for looking backward. Nobody living through the second wave called it that in real time; the label arrived years later, once epidemiologists had the shape of the whole thing to trace. Whatever this one ends up getting called — a fifth wave, a deepening of the fourth, or something researchers haven’t coined yet — the people living inside it don’t get the luxury of waiting for the label to settle before they need care that doesn’t exist.

Jeffrey Steen got out using drug checking, a harm reduction group, a twelve-step program, and a man who happened to check on him at the right hour. That combination worked for him. It is not a treatment system, and it was never meant to be one — it was one person’s improvisation in the absence of anything better, the same improvisation hundreds of thousands of people are running right now with methamphetamine or cocaine as the drug in question. The country has the national data on where stimulants sit in the crisis and a growing body of coverage on where the treatment gap for stimulants specifically is widest — Rize’s newsroom is tracking that same medication shortfall across other substances today, because it isn’t unique to meth and cocaine, just most acute there. What the country doesn’t have, thirty-nine years after it chose sentencing over medicine the last time a stimulant wave broke over it, is a pill. Building one is a research problem. Paying for the one behavioral treatment that already works is not — that’s a budget line, in front of legislators who’ve had the evidence for years and the excuse of “no medication exists yet” for exactly as long. The excuse was already thin. The math just made it indefensible.