From September 1985 to December 1989 a total of 158 patients had primary intracerebral haemorrhage in the population of 116,000 in the Jyväskylä Region, Central Finland. All had the diagnosis confirmed by either computerised tomography or necropsy, and information on cigarette smoking habits was available in 155 patients, 20% of whom were current cigarette smokers. One control was selected for each of the 155 patients matched on sex, age, and residence from the census of Central Finland. The odds ratio of primary intracerebral haemorrhage of current cigarette smokers compared with current non-smokers was estimated on basis of the number of discordant pairs as 1.4 (95% confidence interval 0.7 to 2.8). Adjustment for hypertension or diabetes did not change this estimate. Our data did not show evidence of an positive association of cigarette smoking and risk of primary intracerebral haemorrhage. This assumption was strengthened when the results of previous studies and the present study were pooled giving an odds ratio of 1.0 (95% confidence interval 0.8 to 1.3).