To study the pattern of stroke in Kuwait, prospective data were collected from all patients admitted with first-ever stroke to our hospital over 3 years: 1989, 1992 and 1993. The primary-care registry over the same period was examined to ascertain that all cases of stroke seen by the primary-care doctors were referred to a hospital. Death certificates issued for people who died outside the hospital were also examined. The study included: clinical evaluation, computed tomography, electrocardiography, blood laboratories, serum chemistry and other investigations as required. The overall annual crude incidence rate was 27.6/100,000 population. The age-adjusted annual crude incidence rate was 145.6/100,000 population. No seasonal variation in the stroke incidence was observed in our study. Carotid-territory large infarction represented 46.5% of all strokes followed by intracerebral haemorrhage in 19.9%, lacunar infarction in 17%, basilar infarction in 8.3% and subarachnoid haemorrhage in 1.7%. The overall case fatality rate was 10%. Mortality increased with age and when loss of consciousness was present. The study demonstrated that stroke incidence in Kuwait is very low due to the young average age of the Kuwaiti population.