BACKGROUND AND AIMS: Naloxone is an opioid antagonist used for emergency resuscitation following opioid overdose. Prisoners with a history of heroin injection have a high risk of drug-related death soon after release from prison. The NALoxone InVEstigation (N-ALIVE) pilot trial (ISRCTN34044390) tested feasibility measures for randomized provision of naloxone-on-release (NOR) to eligible prisoners in England. DESIGN: Parallel-group randomized controlled pilot trial. SETTING: English prisons. PARTICIPANTS: A total of 1685 adult heroin injectors, incarcerated for at least 7 days pre-randomization, release due within 3 months and more than 6 months since previous N-ALIVE release. INTERVENTION: Using 1 : 1 minimization, prisoners were randomized to receive on release a pack containing either a single ‘rescue’ injection of naloxone or a control pack with no syringe. MEASUREMENTS: Key feasibility outcomes were tested against prior expectations: on participation (14 English prisons; 2800 prisoners), consent (75% for randomization), returned prisoner self-questionnaires (RPSQs: 207), NOR-carriage (75% in first 4 weeks) and overdose presence (80%). FINDINGS: Prisons (16) and prisoners (1685) were willing to participate [consent rate, 95% confidence interval (CI) = 70–74%]; 218 RPSQs were received; NOR-carriage (95% CI = 63–79%) and overdose presence (95% CI = 75–84%) were as expected. We randomized 842 to NOR and 843 to control during 30 months but stopped early, because only one-third of NOR administrations were to the ex-prisoner. Nine deaths within 12 weeks of release were registered for 1557 randomized participants released before 9 December 2014. CONCLUSIONS: Large randomized trials are feasible with prison populations. Provision of take-home emergency naloxone prior to prison release may be a life-saving interim measure to prevent heroin overdose deaths among ex-prisoners and the wider population.