The suggestion that infection might provoke myocardial infarction prompted a controlled study among 49 patients with myocardial infarction for antibody titres to 22 viruses, and to Mycoplasma pneumoniae and Chlamydia psittaci. The frequency of significant rises in antibody titres was similar in patients with myocardial infarction, 7 (14%) out of 49, and in control patients, 8 (22%) out of 37. Thus, in our limited series we could not confirm the hypothesis that viral or mycoplasmal infection are significantly associated with myocardial infarction. Seven of 15 infections, evenly distributed within the two groups, were due to influenza A virus, which was epidemic in Finland at the time of survey January--March 1980. No acute Coxsackie B virus infections were observed.