This study examined the relationship of employment status and employment-related behaviors to the incidence of coronary heart disease (CHD) in women. Between 1965 and 1967, a psychosocial questionnaire was administered to 350 housewives, 387 working women (women who had been employed outside the home over one-half their adult years), and 580 men participating in the Framingham Heart Study. The respondents were 45 to 64 years of age and were followed for the development of CHD over the ensuing eight years. Regardless of employment status, women reported significantly more symptoms of emotional distress than men. Working women and men were more likely to report Type A behavior, ambitiousness, and marital disagreements than were housewives; working women experienced more job mobility than men, and more daily stress and marital dissatisfaction than housewives or men. Working women did not have significantly higher incidence rates of CHD than housewives (7.8 vs 5.4 per cent, respectively). However, CHD rates were almost twice as great among women holding clerical jobs (10.6 per cent) as compared to housewives. The most significant predictors of CHD among clerical workers were: suppressed hostility, having a nonsupportive boss, and decreased job mobility. CHD rates were higher among working women who had ever married, especially among those who had raised three or more children. Among working women, clerical workers who had children and were married to blue collar workers were a highest risk of developing CHD (21.3 per cent).