Because of women's survival advantage, the impact of myocardial infarction (MI) on long-term mortality in women compared with men may be underestimated. The authors examined this issue in a community sample of 2,462 persons aged > or = 65 years living in New Haven, Connecticut, who were free of MI at baseline and were followed for 10 years (1982-1992). By using proportional hazards models with MI hospitalizations and the sex-MI interaction as time-dependent covariables, survival for the MI cases from the date of MI was compared with survival of persons who, at the same follow-up time, were still alive and free of MI. Women survived longer than men mainly in the absence of MI. The multivariable-adjusted hazard ratios of death were 0.53 in the absence and 0.87 in the presence of MI, and MI was associated with a greater risk of death in women (adjusted hazard ratio = 5.9) than in men (adjusted hazard ratio = 3.6) (p = 0.01 for the sex-MI interaction). When out-of-hospital fatal infarctions were considered, the impact of MI on survival in women compared with men increased. In conclusion, in this elderly cohort, when viewed from a population perspective, MI had a greater impact on mortality in women and significantly narrowed women's typical survival advantage over men.