Consultant physicians and cardiologists were surveyed early in 1987 and 1989 to document the management policies for the treatment of acute myocardial infarction in United Kingdom hospitals and to assess the influence of major clinical trials on these policies. The response rate to both these surveys was high (84% (1178 physicians) in 1987 and 76% (982 physicians) in 1989). The percentage of physicians that reported using antiplatelet therapy "routinely" in acute myocardial infarction rose from 9% in 1987 to 84% in 1989 while those who reported using it "rarely or never" fell from 42% to 3%. Similarly, "routine" use of fibrinolytic therapy rose from 2% to 68%, and use "rarely or never" fell from 53% to 3%. This increase in the reported use of fibrinolytic therapy was accompanied by greater certainty about its efficacy and relative safety and by a general widening of the indications for its use. The use of other treatments in acute myocardial infarction (for example, the general use of anticoagulants, beta blockers, nitrates, calcium antagonists, or prophylactic antiarrhythmic agents) seemed to change little during this period, although the routine use of coronary angiography and oral anticoagulants after fibrinolytic therapy fell substantially between 1987 and 1989 (from 23% to 4%, and from 24% to 7% respectively). Fibrinolytic and antiplatelet therapy were accepted into the routine management of myocardial infarction during a relatively short period that coincided with the reporting of several positive controlled trial results. Clinical trials have rarely been seen to have had such a great impact on practice. In this case the rapid acceptance of the trial results may have been due to the consistency and reliability of the estimates of the size of the benefits (and risks) of therapy seen in these unusually large studies.