The relation of demographic, behavioral, and health-related variables to smoking cessation was studied in 1,178 women and 1,506 men cigarette smokers enrolled in the Framingham Heart Study. Smoking cessation was defined as abstinence from all tobacco products for at least 1 year. Proportional hazard models were used to identify baseline variables associated with long-term cessation. In order to identify factors predictive of cessation in the following 2 years, logistic regression models on person-examination data were used. Recent hospitalization and development of coronary heart disease were predictive of smoking cessation, while diagnosis of cancer or changes in pulmonary function were not. Overall, women were as likely to quit as men; however, women who smoked heavily were the least likely to quit. Secular trends were noted in men only; men were more likely to quit after the release of the 1964 Surgeon General's report on smoking. The relation of illness development and sex to smoking cessation should be considered in developing smoking cessation programs.