There are conflicting reports on the possible positive association between coronary disease and dementia. The objectives of this study were to examine the association between coronary disease, as measured by myocardial infarction and cardiac death, and dementia in a population-based study. By use of the record-linkage system of the Rochester Epidemiology Project, 916 cases of dementia and 916 age (+/-1 year)- and sex-matched controls were identified in Rochester, Minnesota, between 1985 and 1994. From the same population, the authors identified all subjects who experienced a myocardial infarction (defined using standardized criteria) during the period 1979-1998. For myocardial infarction occurring prior to the index year of dementia, the authors used conditional logistic regression (case-control analysis), while for myocardial infarction and death occurring after the index year, they used competing risk survival analysis to account for informative censoring (cohort analysis). Before the index year, the odds ratio for myocardial infarction among cases with dementia compared with controls was 1.00 (95% confidence interval (CI): 0.62, 1.62; p = 1.00). After the index year, patients with dementia had a 46% decreased risk of subsequent myocardial infarction (hazard ratio = 0.54, 95% CI: 0.36, 0.82; p = 0.004) and an 18% decreased risk of cardiac death (hazard ratio = 0.82, 95% CI: 0.70, 0.95; p = 0.010). There was no evidence of a positive association between dementia and preceding myocardial infarction, while there was a decreased risk of myocardial infarction and cardiac death following dementia.