We investigated seasonal variation in the incidence of cerebral stroke among the general population aged greater than or equal to 40 years in November of 1961 in Hisayama, Japan. During the 24-year follow-up period, 311 cases of cerebrovascular diseases occurred. The date or month of onset was determined in 308 cases, of which 51 were classified as intracerebral hemorrhage, 223 as cerebral infarction, and 27 as subarachnoid hemorrhage. We observed a significant seasonality in the incidence of all stroke (p less than 0.01), of intracerebral hemorrhage (p less than 0.05), and of cerebral infarction (p less than 0.01), whereas subarachnoid hemorrhage had no significant seasonal pattern. Subjects less than 64 years of age showed a significant seasonal variation in the incidence of both intracerebral hemorrhage (p less than 0.05) and cerebral infarction (p less than 0.01). A significant seasonal pattern for the incidence of intracerebral hemorrhage was also noted among persons with hypertension (p less than 0.05) or a high serum cholesterol level (p less than 0.05), whereas such a pattern for cerebral infarction was documented among normotensive persons (p less than 0.05) and those with a low serum cholesterol level (p less than 0.01). In addition, the incidences of intracerebral hemorrhage and cerebral infarction were negatively correlated with mean ambient temperature (p less than 0.01 and p less than 0.05, respectively), and all stroke and intracerebral hemorrhage in men were significantly related to intradiurnal temperature change (p less than 0.05 and p less than 0.01, respectively). The significance of the seasonal occurrence of stroke is discussed in relation to relevant risk factors.