Management of stroke patients in specialist stroke units hastens recovery but is not believed to influence mortality. We did a statistical overview of randomised controlled trials reported between 1962 and 1993 in which the management of stroke patients in a specialist unit was compared with that in general wards. We identified 10 trials, 8 of which used a strict randomisation procedure. 1586 stroke patients were included; 766 were allocated to a stroke unit and 820 to general wards. The odds ratio (stroke unit vs general wards) for mortality within the first 4 months (median follow-up 3 months) after the stroke was 0.72 (95% CI 0.56-0.92), consistent with a reduction in mortality of 28% (2p < 0.01). This reduction persisted (odds ratio 0.79, 95% CI 0.63-0.99, 2p < 0.05) when calculated for mortality during the first 12 months. The findings were not significantly altered if the analysis was limited to studies that used a formal randomisation procedure. We conclude that management of stroke patients in a stroke unit is associated with a sustained reduction in mortality.