A novel Canadian harm reduction program allowed 35 individuals with severe alcohol use disorder to substitute pre-rolled cannabis joints for prescribed alcohol doses multiple times daily. Hierarchical mixed-effects modeling revealed a clear substitution effect: participants who used more cannabis consumed less alcohol over the 13-month study period. This represents the first formal evaluation of cannabis integration within managed alcohol programs, which traditionally provide controlled alcohol doses alongside housing support to stabilize drinking patterns and reduce emergency service use. The findings align with emerging evidence that cannabis may serve as a less harmful substitute for alcohol, particularly given cannabis's lower toxicity profile and reduced risk of fatal overdose. However, the small sample size and specialized population—individuals with severe AUD and housing instability—limit generalizability to broader addiction treatment contexts. The intervention's success may also reflect Canada's unique combination of cannabis legalization and established harm reduction infrastructure. While promising for harm reduction advocates, this approach remains experimental and requires larger controlled trials to establish efficacy across diverse populations. The study suggests cannabis substitution could become a valuable tool within comprehensive addiction treatment programs, though regulatory and clinical implementation challenges persist in most jurisdictions.