Melatonin is a candidate neuroprotective drug for ischaemic stroke. Any decision to proceed to clinical trial for such drugs should be based on an unbiased assessment of all available data. Such an assessment should include not only the efficacy of a drug but also the in vivo characteristics and limits--in terms of time window, dose, species and model of ischaemia used--to that efficacy. Here we use a systematic approach to establish the limits to and characteristics of the neuroprotective efficacy of melatonin in experimental stroke. We have used systematic review and meta-analysis to assess the evidence for a protective effect of melatonin in animal models of focal cerebral ischaemia. Fourteen studies were identified describing procedures involving 432 animals. The point estimate for the effect of melatonin was a 42.8% (95% CI 39.3-46.3%) improvement in outcome. Efficacy was greater when ketamine anaesthesia was used, and melatonin was equally effective in permanent or temporary ischaemia. Study quality was generally poor by clinical trial standards, and no evidence was found regarding the efficacy of melatonin in focal cerebral ischaemia in aged, hypertensive or diabetic animals, in species other than rats, or at time windows beyond 2 hr. These findings demonstrate a marked efficacy of melatonin in animal models of focal cerebral ischaemia, identify priority areas for future animal research, and suggest melatonin as a candidate neuroprotective drug for human stroke.