The role of plasma cortisol in the relationship of alcohol consumption and smoking with BP was investigated in a study of 297 Japanese men, aged 50-54 years, who were not receiving antihypertensive agents. They were admitted to the Self-Defense Forces Fukuoka Hospital between January and June 1992 for a detailed pre-retirement health examination. A history of alcohol consumption and cigarette smoking were determined from a self-administered questionnaire. The plasma level of cortisol and BP were determined in the morning of the first admission day. While the plasma level of cortisol was positively related to systolic and diastolic BP, cortisol levels did not vary substantially with alcohol consumption. Both BP and plasma cortisol levels were lower among current smokers than nonsmokers. The lower BP observed among current smokers was ascribed in part (about 20-30%) to the plasma cortisol levels. While the cortisol levels may contribute the lower BP among current smokers, the data did not support its role in mediating the alcohol-BP relationship.