Platelet aggregate ratio (PAR) was measured according to the method of Wu & Hoak in 143 patients after acute myocardial infarction (AMI) and in 54 controls. A PAR < 1 expresses the presence of platelet aggregates. The patients were randomized to aspirin 160 mg/d, or warfarin, or aspirin 75 mg/d + warfarin. In patients on aspirin, PAR was measured 24 h after aspirin intake, and in 76 patients also 2 h after aspirin. The median PAR in patients on warfarin was 0.85, on warfarin + aspirin 0.91 and on aspirin alone 0.94, all significantly lower than the median PAR of 0.97 in the controls. In 14 patients on aspirin the PARs were below a cut-off point of 0.82 (secondary aspirin non-responders). PAR increase significantly 2 h after aspirin intake. In two patients, however, PAR remained low (primary aspirin non-responders). It is concluded that some patients do not seem to respond to aspirin, the clinical implication of which has yet to be determined.