Heroin Basically Disappeared From the Drug Supply. Recovery Culture Hasn’t Caught Up.

Walk into an opioid recovery meeting tonight almost anywhere in the country and you’ll hear people introduce themselves as “a heroin addict.” It’s the word the culture defaulted to — the word 12-step literature was written around, the word treatment intake forms still print, the word a generation of parents and case managers and journalists learned to say first. Here’s the quiet fact underneath that word: most of the people saying it, if you tested what actually put them in that room, never used heroin at all. They used fentanyl, sold to them as heroin, cut with heroin, or not touching heroin in any form. The drug the language is built around has largely left the building.

We are still telling the story of an opioid crisis using the name of a drug most of today’s opioid crisis doesn’t actually contain.

This isn’t a matter of slang lagging behind reality by a news cycle or two. It’s structural. Fentanyl has functionally replaced heroin across most of the U.S. illicit opioid supply — not displaced it at the margins, replaced it, the way a cheaper, more potent, easier-to-synthesize product replaces a harder-to-produce agricultural one whenever the economics allow it. Heroin requires poppy fields, harvest seasons, and transnational smuggling routes vulnerable to interdiction. Fentanyl requires a chemistry set and a supply of precursor chemicals. Once a market fully switches, researchers who study the transition note, there’s little economic logic left pulling traffickers back — heroin is not likely to vanish entirely, but its share of the illegal opioid supply keeps shrinking, market by market, city by city.

The transition nobody voted on

This has happened before, which is the part worth sitting with. In the mid-2000s, the U.S. opioid crisis ran on prescription pills — OxyContin, Vicodin, diverted from pharmacies and medicine cabinets. When the DEA and manufacturers tightened pill supply starting around 2010, people already physically dependent didn’t stop being dependent; they moved to the cheaper, more available substitute, which at the time was heroin. That transition — pills to heroin — is the one most addiction-recovery language and imagery was actually built to describe: a person who started with a legitimate prescription and ended up buying dope on a corner. Then, starting in earnest around 2013 in the Northeast and spreading west over the following decade, fentanyl began cutting into and then overwhelming the heroin supply itself, because it was cheaper to produce, easier to smuggle in small, potent quantities, and more profitable per dose. Each transition happened in the drug supply first and in public understanding years later. We are living in the second lag right now, and recovery culture — the meetings, the media, the treatment paperwork — is still narrating the first one.

The consequence isn’t just semantic. A treatment intake form that asks “heroin use: yes/no” is asking the wrong diagnostic question to a population that mostly can’t answer it accurately, because most street fentanyl in the U.S. isn’t sold honestly as fentanyl — it’s sold as heroin, as counterfeit oxycodone, as “dope,” under names that tell the buyer nothing true about what’s actually in the bag. Someone can be fentanyl-dependent, fentanyl-tolerant, and at fentanyl-specific overdose risk while genuinely, honestly believing they use heroin, because that’s the word the person who sold it to them used. Clinical protocols built around heroin’s pharmacology — its half-life, its withdrawal timeline, its response to standard naloxone dosing — don’t map cleanly onto fentanyl, which clears differently, binds differently, and increasingly resists a single dose of the drug that used to reliably reverse an overdose.

Why the naloxone playbook is straining

That resistance is showing up in the numbers behind what’s otherwise genuinely good news. The country is in the middle of a real, sustained decline in overdose deaths — falling fentanyl potency is doing real work in that decline, alongside years of accumulated naloxone access. But researchers tracking the street supply are watching a new layer of adulterants — medetomidine, xylazine, and other non-opioid sedatives — move into the space heroin used to occupy, and several of them don’t respond to naloxone at all, because naloxone only reverses opioids and these newer additions frequently aren’t opioids. Every transition in this crisis has followed the same shape: the response system calibrates to the drug currently causing the most damage, and the supply moves again before the calibration finishes.

That resistance is showing up in the numbers behind what’s otherwise genuinely good news.

If you’re the person in the room calling yourself a heroin addict because that’s the only word anyone gave you, none of this is a correction aimed at you — it’s a correction aimed at the systems that handed you a word that was already out of date. What you survived doesn’t need the right pharmacological name to count. But the people building the next round of treatment protocols, harm reduction supplies, and intake forms need to stop asking a 2010 question about a 2026 drug supply, because the gap between the word and the substance is exactly where naloxone fails to work and where a clinician miscalibrates a taper.

Some of this is already changing where it counts most: newer harm reduction programs have largely stopped asking “do you use heroin” and started asking what color the powder was, whether a test strip was used, whether the person has ever needed more than one dose of naloxone to come back — questions built for the drug that’s actually circulating, not the one on the form. That shift matters because it changes what gets handed to someone at the door: multiple naloxone doses instead of one, fentanyl test strips instead of a pamphlet, a conversation about xylazine wound care that a heroin-only script would never think to include. Peer support workers who use drugs themselves, or who did, have generally led this shift faster than clinical institutions have, for the obvious reason that they’re the ones who noticed the bag stopped matching the word first.

The slower-moving piece is language itself, and language is not a small thing in recovery work — it’s how people find each other, how a meeting decides who belongs in the room, how a family member describes what happened to a sibling. “Heroin addict” carries decades of cultural weight, some of it stigmatizing, some of it a hard-won identity people built real recovery on top of. Nobody is well served by simply deleting the word. What’s needed is smaller and more honest: an acknowledgment, spoken out loud in the rooms where it matters, that the word and the substance parted ways years ago, and that the gap between them is where the next preventable overdose is most likely to happen — not because anyone lied, but because the drug supply moves faster than the vocabulary built to describe it.

Heroin isn’t gone from American culture — it’s still the word, still on the forms, still in the meetings. It’s increasingly just not in the bag. That gap between the name and the substance is where the next preventable overdose is most likely to happen, and closing it is cheaper than almost anything else discussed in this crisis: it costs a rewritten intake form, an updated naloxone-training script, and a willingness to say, out loud, that the drug we built this entire vocabulary around mostly isn’t the drug in front of us anymore. Learn what’s actually in today’s opioid supply and how to respond to an overdose →