A prospective, randomized trial was designed to assess the efficacy of intracoronary thrombolytic therapy with streptokinase (STK) in acute myocardial infarction. Sixty-four patients with acute myocardial infarction were randomized within 6 hours of onset of symptoms to 1 of 3 groups. Sixteen patients were treated by conventional means (control group). Nineteen patients underwent coronary arteriography and received corticosteroids and intracoronary and intravenous nitroglycerin (NTG group). Twenty-nine patients received management identical to that of the NTG group, with the addition of intracoronary STK therapy (STK group). Recanalization was demonstrated in 21 of 29 patients (72%) in the STK group. Global and regional ejection fraction (EF) was determined by radionuclide ventriculography before any intervention and 7 to 10 days later. No significant improvement in global EF was achieved in the control and NTG groups. In STK patients as a group, global EF did not increase significantly; however, in patients recanalized with STK, EF improved from 42 +/- 17% to 49 +/- 16% (p = 0.023). All groups showed wide variability of response. Improvement in global EF of more than 5% was noted in 44% of patients recanalized with STK. When subgrouped on the basis of initial global EF of 45% or less or more than 45%, only patients recanalized with STK with an initial EF of 45% or less had an improved global EF (from 30 +/- 10% to 42 +/- 10%, p = 0.015).(ABSTRACT TRUNCATED AT 250 WORDS)