The reliability of serum MM3/MM1 ratio as a marker for AMI was evaluated in 81 consecutive coronary-care admission. Of these, 62 patients were diagnosed as definite infarction and 19 patients as angina, using WHO criteria. The MM3/MM1 ratio in the admission serum sample was 1.85 +/- 2.01 in 62 patients with definite infarction; in contrast, the MM3/MM1 ratio was 0.24 +/- 0.14 in 19 patients with angina (P less than 0.01 for MI group vs angina group). Among the AMI patients admitted within 12 h after onset, the diagnostic accuracy of the ratio MM3/MM1 was higher than CK, CK-MB, LDH, LD1/LD2 and GOT and similar to Mb in the initial serum samples. It seems that the MM3/MM1 ratio was a better enzyme marker than the others in the early stage of AMI, especially before 12 h.