Serum levels of ventricular myosin heavy chains were quantitated in patients with acute myocardial infarction using a competitive radioimmunoassay involving monoclonal antibodies to the b-type myosin heavy chains of a human ventricle. Among the seven antibodies selected for their higher affinity for ventricular myosin heavy chains, only four antibodies detected significant and variable myosin amounts in the serum samples of nineteen patients with acute myocardial infarction; the same antibodies occasionally detected, if at all, low myosin amounts in the sera of patients with no clinical sign of myocardial infarction, and no myosin in the serum of the healthy control subjects. The peak levels of myosin release were observed 4.6 +/- 0.5 days (n = 13, P less than 0.01) after myocardial infarction and correlated rather well with the measured creatine kinase peak levels (the correlation coefficients were between 0.75 and 0.81, P less than 0.01, depending on the monoclonal antibody used for myosin determination). The time courses of myosin release varied according to the complexity of the heart attack observed. It was concluded that the titration of serum myosin was probably of little clinical value for therapeutic intervention during the acute phase of myocardial infarction; it could, however be an effective tool for retroactive detection of an infarct and for late estimation of infarct size.