In a nine-year follow-up of a southern California community of 2,852 men and women aged 60-79 years, systolic blood pressure was a significant predictor of subsequent cancer mortality in men. This effect was independent of age, antihypertensive medication, smoking, obesity, and plasma cholesterol. Trends in women were similar but not statistically significant. Compared with those still alive, higher initial systolic blood pressure levels were apparent in those who died of colon cancer, stomach cancer, and all other cancers combined except for lung and prostate cancer. Possible mechanisms for this association and the implications of the data with regard to the benefits of measures to treat high blood pressure or lower population distribution of blood pressure are discussed.