The Ontario government recently said it will cut provincial funding for seven supervised drug consumption sites in Toronto, Ottawa, Niagara, Peterborough and London, with 90 days given to wind down their operations.In their place, the province is spending $378 million on 19 Homelessness and Addiction Recovery Treatment (HART) hubs, which explicitly exclude supervised consumption and needle exchange services.While Premier Doug Ford’s government’s framing of supervised consumption as a “failed experiment” is selective, it’s not baseless in the way that defenders of these sites sometimes imply. By certain measures, the sites have not lived up to their potential: the program has not clearly reduced provincewide overdose deaths, and the communities that host them bear costs that defenders too often dismiss.But “failure” requires a definition of success, and the government’s is not the only one that matters. HART hubs offer care for people on a recovery pathway, while supervised consumption sites exist for those who are not on that pathway yet or who have left it. Entirely replacing safe consumption sites with HART hubs doesn’t address the primary function these sites have served: keeping people who are not in treatment alive. By this measure, the sites are a clear success. Success by whose measure?Whether supervised consumption sites “succeed” depends entirely on what they’re expected to do. If we judge the sites against Health Canada’s stated goals — keeping people alive on site, connecting them to treatment, reducing infections and lessening strain on emergency services — the evidence is strong. No one has ever died of an overdose inside a supervised consumption site in Canada. And sites across the country have reversed more than 50,000 overdoses since 2017. In Ontario, where more than 2,200 people died from opioid toxicity in 2024 and fentanyl was involved in more than 83 per cent of those deaths, the sites function as a last line of defence for people at highest risk. Since 2021, about one in five opioid toxicity deaths in Ontario has occurred among people experiencing homelessness — the same population these sites primarily serve. Beyond lives saved, safe consumption sites generate measurable returns the government’s own cost-benefit logic should recognize: Vancouver’s Insite refers thousands of clients to health and social care monthly and a Calgary cost analysis found each overdose managed at a supervised consumption site saved approximately $1,600 in avoided ambulance and emergency department costs. Those resource savings are especially important amid severe emergency room overcrowding in Ontario. Where the sites fall shortBut the Ontario government’s claim that these programs lack population-level impact is not just rhetoric: the two largest provincial-level studies — covering British Columbia and Ontario — found no statistically significant effect on opioid mortality, emergency department visits or hospitalizations. A systematic review also found that the studies carried high risk of bias because they didn’t account for confounding factors like housing, treatment access and drug supply composition. A neighbourhood-level study of Toronto found a two-thirds reduction in overdose mortality within 500 metres of sites, but that finding did not replicate at larger geographic scales. One possible explanation for this lack of effect is coverage: Ontario’s supervised consumption sites provided roughly 150 spaces accommodating up to 9,000 episodes per day, while the province may have 300,000 to 400,000 at-risk opioid users. Expecting a handful of supervised consumption sites to reduce provincewide overdose mortality is akin to stationing a single fire truck in a forest and asking why the wildfire kept burning.Community concerns, which the Ford government often cites in arguing that safe consumption sites have failed, are similarly grounded in real experiences. A study published in JAMA Network Open examining Toronto’s safe consumption sites found no long-term rise in overall crime, as well as fewer assaults and robberies. But the same study documented initial increases in break-and-enters near a number of sites. Read more: New study: Some crimes increased, others decreased around Toronto supervised consumption sites Nearby residents have also described visible disorder, open drug use and discarded equipment. Dismissing these experiences as NIMBYism or overblown only undermines the case for safe consumption sites. Advocates need to take these concerns seriously if they want the sites to survive politically.What closures get wrongDespite this mixed evidence, there’s little to suggest that Ontario’s plan to shift patients to the HART hubs will lead to success. The Ontario government has cited a Canadian Centre of Recovery Excellence study that found no increase in mortality after the closure of one overdose prevention site in Red Deer, Alta. But that study’s authors have acknowledged it was inconclusive because it covered only a six-month period. And a single study of a single site closure does not constitute an evidence base for dismantling an entire network of services across a province where opioid deaths remain catastrophically high.Supervised consumption sites are not beyond criticism: they can be better designed, better integrated and more responsive to the communities that host them. But improving them requires better policy, not selective evidence and site closures. The $378 million committed to HART hubs could expand addiction treatment without eliminating the services that keep people alive. Adding treatment capacity does not require removing the safety net beneath it.Daniel Eisenkraft Klein does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.