Recent studies suggest that plasminogen activator inhibitor (PAI-1) may be a risk factor for recurrent myocardial infarction. We measured PAI-1 activity and antigen and tissue-type plasminogen activator (t-PA) antigen in 35 (20 nondiabetic and 15 diabetic) subjects with no clinical or electrocardiographic evidence of ischemic heart disease and in 74 (50 nondiabetic and 24 diabetic subjects) who had survived a myocardial infarction in the preceding 6-24 months. Levels of PAI-1 activity (18.7 +/- 5.6 versus 12.0 +/- 3.8 arbitrary units [AU] per milliliter, p = 0.001) and t-PA antigen (7.0 +/- 1.9 versus 4.6 +/- 2.0 ng/mL, p = 0.001) were significantly higher in diabetic compared with nondiabetic control subjects. Survivors of myocardial infarction had higher levels of PAI-1 activity and antigen and t-PA antigen than control subjects, and the diabetic survivors had higher levels of PAI-1 activity (25.3 +/- 6.7 versus 20.1 +/- 7.1 AU/mL, p = 0.004) and t-PA antigen (10.6 +/- 4.3 versus 8.4 +/- 3.3 ng/mL, p = 0.03) than the nondiabetic survivors. No difference in PAI-1 antigen levels was found between the diabetic subjects and either the nondiabetic control subjects or survivors of myocardial infarction. After venous occlusion in control subjects, there was a significant increase in PAI-1 antigen (mean 26.7%, range 14.1-58.1% in nondiabetics and mean 25.2%, range 6.2-39.7% in diabetics) and t-PA antigen (mean 78.3%, range 13.6-186.2% for nondiabetic and mean 40.7%, range 17.5-76.2% for diabetic subjects), but in the survivors of myocardial infarction, no significant effect of venous occlusion was observed.(ABSTRACT TRUNCATED AT 250 WORDS)