Little is known about incorporation of new knowledge from randomised clinical trials into clinical practice. Thrombolytic therapy was shown to reduce the mortality of acute myocardial infarction in several large trials published during 1986-88. To examine the effect of these data on clinical practice, we analysed the supply of thrombolytic drugs in a representative English region (population 4.7 million) in 1987-92. During the study period there were over 10,000 hospital admissions per year in the region for acute myocardial infarction. From a very low initial level, thrombolytic drug use rose slowly for several years after publication of the trial results and reached a plateau in 1991-92. Rates of use per 1000 patients admitted with myocardial infarction varied almost six-fold between districts in 1989-90 and over two-fold in 1991-92. Level of use attained by districts in the latter period was strongly associated with the extent of their previous participation in multicentre trials of thrombolysis (p = 0.003); we estimate that 35-50% of patients admitted with acute myocardial infarction were receiving thrombolytics. The full potential of thrombolytic treatment has still not been achieved in routine care and the limiting factors need to be defined.