Despite the proven benefits of thrombolytic therapy in acute myocardial infarction, concern for its complications, especially in patients misdiagnosed with myocardial infarction, has led to hesitancy in its use. Historical, clinical and electrocardiographic criteria were developed for enrolling patients with suspected acute myocardial infarction into thrombolytic trials by noncardiovascular specialists. The incidence of misdiagnosis of myocardial infarction and the clinical outcomes when these criteria were used were evaluated for 1,387 consecutive patients given thrombolytic therapy. Twenty-five community hospitals and 7 interventional centers were the sites of enrollment. Most patients (63%) were enrolled from community hospitals. Criteria for thrombolytic therapy included: symptoms of acute myocardial infarction < 6 hours but > 20 minutes, and not relieved by nitroglycerin; and ST-segment elevation > or = 1 mm in 2 contiguous leads or ST-segment depression of posterior myocardial infarction. Exclusion criteria reflecting increased risk of bleeding were used. A final diagnosis of myocardial infarction was based on creatinine kinase-MB, electrocardiographic and ventriculographic evaluation. Acute myocardial infarction was misdiagnosed in 20 patients (1.4%; 95% confidence interval 0.8-2.0%). These patients were demographically similar to those with acute myocardial infarction. All misdiagnosed patients survived; no significant adverse events occurred. Thus, in several clinical settings, a simple algorithm with specific criteria was used for diagnosing acute myocardial infarction and administering thrombolytic therapy. The inclusion criteria used in this study led to a low rate of misdiagnosis.