STUDY OBJECTIVE--To examine the effect on mortality of stopping smoking after myocardial infarction and the psychosocial factors that influence the decision to stop. DESIGN--Analysis of smokers in a large prospective study. Self completed questionnaires provided information on psychosocial factors. SETTING--Coronary care units at six English hospitals participating in a multicentre clinical trial. SUBJECTS--These comprised consenting myocardial infarction survivors who had been identified as smokers and who completed questionnaires within seven days of infarct at six hospitals participating in the Anglo-Scandinavian study of early thrombolysis. The 532 patients identified have been followed for over five and a half years. The main outcome measure was five year all cause mortality. MAIN RESULTS--Smokers who stopped within one month showed significantly reduced mortality compared with those who persisted, adjusting for other prognostic indicators (odds ratio 0.56, 95% confidence interval 0.33, 0.98). Overall, 74% stopped smoking. Being married, low life stress levels before infarct, and higher social class were associated with stopping smoking but the differentials were small. Of the clinical variables, a final diagnosis of definite myocardial infarction was associated with stopping smoking. All associations remained after multiple logistic regression. CONCLUSION--Smoking cessation can halve the smokers' odds of dying after myocardial infarction and psychosocial factors play a small but important role in the important decision to stop smoking. Health professionals should continue to stress the importance of stopping smoking to all patients as there is little evidence to support specific directing of advice to relatively "stress or "socially isolated" groups.