The immediate and delayed effects of urokinase and heparin on minimal cross-sectional area of a patent ischemia-producing coronary artery were prospectively investigated in 43 patients with unstable angina. After baseline angiography, patients were randomized to 3 different treatment groups: group I--urokinase (1,000,000 U intravenous bolus dose), followed by heparin infusion 3 hours later; group II--heparin (10,000 U intravenous bolus, followed by continuous infusion); and group III--conventional therapy only (intravenous nitroglycerin, beta blockers and calcium antagonists). Angiography was repeated at 1 hour and at 8 days of treatment and minimal cross-sectional area was determined in the 35 patients who completed the study. In group I, minimal cross-sectional area increased from 0.84 +/- 0.48 mm2 at baseline to 0.94 +/- 0.49 mm2 at 1 hour (p less than 0.05), and to 1.00 +/- 0.51 mm2 at 8 days (p less than 0.01 vs baseline). In group II, a significant increase in minimal cross-sectional area was observed only at the 8-day angiography (0.64 +/- 0.39 mm2 at baseline; 0.67 +/- 0.37 mm2 at 1 hour [p = not significant]; and 0.79 +/- 0.48 mm2 at 8 days [p less than 0.01] vs baseline). In group III, no significant changes in minimal cross-sectional area occurred either at 1 hour or at 8 days. Thus, both urokinase and heparin improved lesion geometry in patients with unstable angina, although a large individual variation was noticed. The effect occurred earlier with urokinase than with heparin.