Smoking patterns of 205 male patients were examined over a period of eight years after a primary myocardial infarction. Smoking data from their wives at one year after the heart attack were also examined. A marked, persistent reduction in smoking was found among the men. Smoking patterns of wives remained essentially unchanged. Smoking patterns before the heart attack were not related to demographic variables, except for the associated between smoker-nonsmoker status and social measures. Conceptions of susceptibility, threat, and power of prevention drawn from theoretical models on preventive health behavior were employed for analysis. High proportions of husbands and wives reported belief in smoking as important in the etiology of the heart attack, and in possibilities of prevention. Specific beliefs concerning threat, susceptibility, and prevention were not found to be related to the massive drop in smoking behavior among the men nor to stability in pattern among the wives. Alternative explanations are reviewed.