The improvement in survival in patients undergoing thrombolytic therapy in myocardial infarction is determined by the delay between coronary occlusion and reperfusion. The REPerfusion in Acute Infarction Rotterdam (REPAIR) study was designed to examine the feasibility and safety of prehospital thrombolysis with alteplase (rt-PA, 'Actilyse'). A small portable ECG computer system is used to confirm the presence of a large myocardial infarction (at least 1.0 mV ST-deviation) 'on the spot'. Between 22 June 1988 and 1 January 1991, 226 patients were treated by the ambulance service after the evaluation of 9052 patients complaining of chest pain. Therapy could be initiated within an average of 100 +/- 56 min (SD) after the onset of symptoms, and within 22 +/- 9 min after ambulance arrival. Three patients were defibrillated during transportation. Six patients (3%) died after arrival in the hospital. The time gained by prehospital treatment was 47 min (95% confidence limits 44-51 min) in comparison with 220 patients who did not meet the criteria for prehospital thrombolysis, but received thrombolytic therapy as soon as possible after hospital admission. The developed procedure allows rapid and safe initiation of thrombolytic therapy in selected patients, even in the absence of a physician. The observed low mortality supports the concept that prehospital thrombolytic therapy is indeed beneficial to the patient.