A series of 200 consecutive patients with autopsy-proven acute myocardial infarction (AMI) was retrospectively studied in order to assess the degree of clinico-pathological agreement and to detect the reasons for disagreement. A correct clinical diagnosis of AMI was made in 86 cases (Group A = 43%) and was missed in 114 cases (Group B = 57%). Atypical presentation and concealed history were more common in group B. The AMI qualified to be the main disease in 83 patients of group A and in 81 of group B and was considered a contributory cause of death in three of group A and in 33 of group B (P less than 0.01). The mean number of diseases coexisting with the main disease for each patient was significantly lower in group A than in group B (P less than 0.01). The mean age was 65.2 +/- 12 years for group A patients and 69.1 +/- 12 years for group B patients (P less than 0.02). With the patients grouped according to age (group I: less than 60 years = 46 cases; group II: greater than or equal to 60 years = 154 cases), the diagnostic accuracy was 61% in group I and 38% in group II (P less than 0.01). Groups I and II did not differ in clinical presentation, ECG and enzyme diagnostic accuracy, while the number of diseases coexisting with the AMI was significantly higher in group II (P less than 0.001). Ageing, the atypical presentation and the coexistence of several diseases seem to account for most of the unrecognized AMI.