For clinical prognosis, LV function is best described in terms of ejection fraction (EF) or end-systolic volume (ESV). Because deterioration of LV function is the most important adverse prognostic factor for patients after recovery from myocardial infarction, any agent which improves function should improve long-term survival. Clinical trials in which LV function has been measured have shown that early administration of all the presently available thrombolytic drugs improves LV function. There is, however, no evidence that one drug is better than another for preservation of LV function. Indeed in our double-blind comparison of streptokinase (SK) with tissue plasminogen activator (tPA) in which intervention was made at 2.5 +/- 0.6 h after onset, EF at 3 weeks after infarction was 58 +/- 12% for SK (n = 116) vs 58 +/- 12% for tPA (n = 124). The 95% confidence limit for the difference was 0 +/- 3%. Patency rates on the infarct-related artery at 3 weeks were similar in the two groups (75% SK, 76% tPA). Besides reflecting infarct size, LV function is also influenced by healing of the infarct, infarct expansion and LV remodelling. Two clinical trials have shown that captopril treatment started at 1-4 weeks after infarction and continuing for 1 year can prevent progressive LV dilatation in patients who have not received thrombolysis. One possible explanation for our own finding of similar effects of SK and tPA on LV function at three weeks after infarction, despite presumably more rapid reperfusion in patients given tPA, is that later reperfusion with SK prevented infarct expansion.(ABSTRACT TRUNCATED AT 250 WORDS)