Homelessness is a major public health and social issue, and research on the effectiveness of housing and harm reduction interventions provides fragmented findings that are difficult to compare across studies. The aim of this systematic review was to assess the effects of housing programmes (including Housing First) and alcohol harm reduction interventions among adults experiencing homelessness. Six databases (PubMed, Embase, Scopus, Web of Science, PsycINFO, and the Cochrane Library) were searched, identifying 1385 records. After screening, 159 publications were assessed in full text, of which nine met the inclusion criteria (total n = 6039 participants). Most studies were randomised controlled trials, with follow-up periods ranging from 12 to 84 months.

Housing interventions consistently increased housing stability and reduced returns to homelessness. Alcohol harm reduction programmes (HaRT-A and HaRT-A + XR-NTX) were associated with reductions in alcohol consumption, alcohol-related harms, and improvements in quality of life. Mental health-related outcomes, including symptom severity, functioning, recovery-oriented outcomes, and quality of life, showed heterogeneous and often non-significant effects across studies. This heterogeneity reflected variability in outcome constructs, measurement instruments, and study designs, limiting comparability across studies. The included studies primarily involved high-need populations characterised by severe mental health conditions, substance use disorders, and high service use, and samples were predominantly male. These populations are not representative of the broader population of adults experiencing homelessness, limiting the generalisability of the findings across sex and gender groups. This review synthesises recent controlled evidence published between 2019 and 2025 and is framed as an update systematic review intended to complement, rather than replace, the substantial pre-2019 evidence base. Restricting the review to recent controlled studies may introduce temporal selection bias, including underrepresentation of foundational efficacy trials and potential overrepresentation of follow-up analyses or context-specific implementations. Accordingly, the findings should be interpreted in conjunction with earlier foundational trials.