Minnesota's Opioid Deaths Fell for a Third Straight Year. Its Total Overdose Toll Still Rose.

Minnesota's opioid death toll has now fallen three years running. The state health department's preliminary count for 2025 is 671 opioid-involved overdose deaths, down from a peak of 1,039 in 2022 — a 35 percent decline in three years, and the kind of number that gets read aloud at press conferences.

The number one line below it in the same report undoes the mood. Total drug overdose deaths in Minnesota rose 3 percent in 2025, from 994 to 1,025. The state did not lose ground on opioids and then hold it. It lost ground on opioids while losing more ground somewhere else.

That somewhere else is stimulants. Fatal overdoses involving psychostimulants such as methamphetamine rose 11 percent between 2024 and 2025, and fatal overdoses involving cocaine rose 6 percent, according to the Department of Health's Statewide Trends in Drug Overdose: 2025 Data Update, released Oct. 27. The report puts the shape of the problem in one sentence of its summary: "As opioid-involved overdoses decrease and stimulant-involved overdoses increase, Minnesota is entering a new phase that requires distinct responses to individual substances and mixed drug use."

What the 2025 numbers actually show

The opioid side of the ledger improved almost everywhere it was measured. Fentanyl-involved deaths fell 6 percent, from 602 to 564. Hospital-treated nonfatal opioid overdoses slipped 2 percent, from 3,685 to 3,617 visits. Within that flat total, the geography split: the Twin Cities metro was essentially unchanged at 2,983 to 2,984 visits, while Greater Minnesota fell 10 percent, from 702 to 633.

The composition of those nonfatal opioid cases is shifting faster than the total. Suspected synthetic-opioid overdoses — fentanyl and its relatives — rose 17 percent, from 1,343 to 1,573 hospital visits, while heroin-involved visits collapsed 44 percent, from 150 to 84. The department is careful about what that means: the rise in identified fentanyl cases, alongside a 12 percent drop in visits coded as "other or unspecified" opioids, "may reflect improved identification of fentanyl over time." In other words, some of the increase in the fentanyl column is a sorting change, not an exposure change.

Then the stimulant columns, which are not ambiguous in the same way. Nonfatal stimulant overdoses rose 11 percent overall, from 1,047 to 1,162 hospital visits, with methamphetamine up 17 percent (427 to 499) and cocaine up 7 percent (352 to 376). The metro drove the meth increase, up 22 percent from 315 to 383 visits. Greater Minnesota saw cocaine visits rise 19 percent, from 43 to 51.

Even after that increase, the 2025 stimulant total of 1,162 sits below every year from 2019 to 2023, when annual visits ran between 1,170 and 1,368. That is not a consolation. It means the 2025 rise is a partial recovery toward a baseline the state never actually escaped, at the same moment its opioid numbers finally started to look like progress.

Fewer than one in five states is reporting an increase

The detail that reframes the whole report is buried on page two. Nationally, overdose deaths are expected to fall for a third consecutive year. Fewer than 20 percent of states are reporting increases in 2025 preliminary data — "and all of these states are located in the upper Midwest and Southwest regions."

Minnesota is one of a small minority of states moving the wrong direction, and it is moving that way in the company of its neighbors. That pattern is worth more attention than it has received. If the national decline is being driven by the shrinking fentanyl supply reaching the street, or by naloxone saturation, or by sustained treatment expansion, those forces should not stop at the state line. A regional cluster of increases suggests something local to the upper Midwest supply — and the report's own data points at what.

Mixed drug use is the default, not the exception

Seventy percent of Minnesota's overdose deaths since 2020 have involved more than one substance. That is not a rounding artifact; it is the modal death.

The specific pairing driving it is opioid plus stimulant. Deaths involving both have more than doubled since 2019, from 154 to 371, and in 2025 they accounted for 61 percent of all multi-substance deaths. Over the same six years, deaths involving neither a stimulant nor an opioid fell 46 percent, from 232 to 125.

The hospital lab data show the same thing from underneath. Minnesota's surveillance system tests samples from a subset of emergency department patients, and in 2025 it found that among opioid-positive cases, 65 percent also had amphetamines present and 33 percent also had cocaine. Sedatives such as xylazine and medetomidine appeared in 17 percent of samples — every one of them co-detected with fentanyl.

That last figure is the one to sit with. The animal sedative that has reshaped the eastern U.S. heroin supply is present in Minnesota, but only as a passenger on fentanyl, never on its own. A person buying a bag of white powder in Minneapolis is not choosing between an opioid and a sedative. They are getting a mixture whose composition nobody at the point of sale has verified.

Why the stimulant side is harder to treat

There is no naloxone for methamphetamine. That single fact explains much of the asymmetry between the two halves of this report, and it is the reason the state's opioid infrastructure — naloxone distribution, MOUD, peer recovery, syringe services — cannot simply be pointed at the new numbers.

"A key challenge is that there are fewer tools to treat stimulant use and related harms," the report states, and the treatment guidance it offers healthcare providers is correspondingly modest: recognize stimulant overdose symptoms such as chest pain, overheating, severe agitation, seizures or stroke-like presentations; screen for substance use disorder; consider off-label medication options from the ASAM/AAAP clinical practice guideline on stimulant use disorder; and — for the opioid half of a mixed overdose — offer or arrange medication for opioid use disorder.

What the guideline can and cannot offer

The off-label caveat is doing real work in that sentence. The 2024 ASAM/AAAP guideline is the field's most complete attempt to organize what can be offered for stimulant use disorder, but it is organizing a thin evidence base: contingency management, the structured-incentive behavioral intervention, carries the strongest trial evidence, and the medication options are borrowed and modest. Compare that to opioid use disorder, where buprenorphine and methadone have decades of mortality data behind them.

The report also repeats a message that runs against consumer intuition about harm reduction: fentanyl may be present in stimulants, so naloxone should still be given in a suspected mixed overdose, and test strips are worth distributing to people who believe they are using only meth.

The alert system that launched four weeks before the report

The one genuinely new piece of infrastructure in the report is MN-OD-ALERTS, a statewide overdose alert system the Department of Health launched Oct. 1 in partnership with the Office of Emergency Medical Services, funded by the Legislature after a proposal from Gov. Tim Walz.

It runs on EMS data. When the system flags a potential opioid spike, health department staff pull patient-level detail and cross-check it against hospital visits and reports from community partners to decide whether the signal is a real county-level change. If it is, local and Tribal public health agencies are notified and can activate a response using the department's community response guide. Seventy-six of Minnesota's 87 counties have supplied a contact.

Dylan Ferguson, director of the Office of Emergency Medical Services, framed the logic in the release: EMS crews "are often the first to see when something is changing in a community," and using their data to detect trends "gives public health and community partners an opportunity to act earlier, before a concerning trend becomes a larger crisis."

A tool tuned to the improving half of the crisis

The design has an obvious limit, and it is the same limit that shows up in every overdose early-warning system built on 911 calls. It detects what already generated an emergency response. A cluster of fatal overdoses in which nobody called — the case where a user is alone, or where the people present are afraid to call — will not trip the alert. Neither will a slow rise in stimulant deaths, because the system as described tracks opioid spikes. The state now has a near-real-time instrument tuned to the half of the crisis that is improving.

What the report counts, and what it doesn't

Two measurement caveats sit inside the document and deserve to travel with the numbers.

The first is a floor problem. The report counts only hospital-treated nonfatal overdoses. It explicitly excludes naloxone reversals in the community and overdoses EMS treats without transporting the patient — which, in a state that has spent years pushing naloxone into libraries, shelters and schools, is a growing share of all reversals. The nonfatal side of the ledger is therefore not a measure of how often people overdose. It is a measure of how often an overdose got far enough to reach a hospital bed.

The second is an identification problem. Fatal overdoses are classified by medical examiner toxicology. Nonfatal ones are classified by what the patient reports or what their symptoms suggest, because toxicology is rarely ordered for clinical care. Methamphetamine identified by a urine screen and methamphetamine identified by a patient's account are not the same measurement, and the report says so. Its own surveillance data underlines the gap: in an earlier analysis of 2020 and 2021 cases, only 52 percent of patients with opioids detected in toxicology had a diagnosis related to opioid use, and only 57 percent of those with stimulants detected had a stimulant-related diagnosis.

The treatment question the state now faces

Minnesota has built a genuinely effective opioid response. The report credits expanded naloxone access, MOUD in clinics, jails and Tribal health programs, peer recovery services and syringe service programs, funded through the Comprehensive Drug Overdose and Morbidity Prevention Act and opioid settlement dollars. The three-year decline is the return on that.

The difficulty is that the same infrastructure has no equivalent for the substance now killing more people. A state can hand out naloxone at a library, but it cannot hand out a methamphetamine antagonist, because none exists. What it can do is make sure the opioid addiction treatment system it built is screening for stimulant use alongside opioid use, that contingency management is actually available rather than merely recommended, and that people presenting with a methamphetamine problem are not bounced out of a system designed around a different drug.

The report's own prevention guidance for funders and policymakers asks for exactly that: sustain the opioid investments, and expand prevention and treatment of stimulant use disorder. One line, at the bottom of page eight, in a document whose headline will be the decline.