Aims Release from prison is a high‐risk period for mortality. We examined the impact of opioid substitution therapy (OST), for opioid dependence during and after incarceration, upon mortality post‐release. Design A cohort was formed of all opioid‐dependent people who entered OST between 1985 and 2010 and who, following first OST entry, were released from prison at least once between 2000 and 2012. We linked data on OST history, court and prison records and deaths. Setting New South Wales (NSW), Australia. Participants A total of 16 453 people released from prison 60 161 times. Measurements Crude mortality rates (CMRs) were calculated according to OST retention; multivariable Cox regressions for post‐release periods were undertaken to examine the association between OST exposure (a time‐dependent variable) and mortality post‐release, for which covariates were updated per‐release. Findings There were 100 978 person‐years (PY) post‐release; 1050 deaths occurred. Most received OST while incarcerated (76.5%); individuals were receiving OST in 51% of releases. Lowest post‐release mortality was among those continuously retained in OST post‐release CMR 4 weeks post‐release = 6.4 per 1000 PY; 95% confidence interval (CI) = 5.2, 7.8, highest among those with no OST (CMR = 36.7 per 1000 PY; 95% CI = 28.8, 45.9). Multi‐factorial models showed OST exposure in the 4 weeks post‐release reduced hazard of death by 75% (adjusted hazard ratio 0.25; 95% CI = 0.12, 0.53); OST receipt in prison had a short‐term protective effect that decayed quickly across time. Conclusion In New South Wales, Australia, opioid substitution therapy in prison and post‐release appears to reduce mortality risk in the immediate post‐release period.