A study in 1995 at Harstad District Hospital concluded that too few patients with acute myocardial infarction had received thrombolytic treatment and that the in-hospital delay before administration of thrombolytics had been too long. To evaluate the effect of improvements in medical and nursing routines, data on all patients with acute myocardial infarction treated between October 1996 and October 1997 (n = 122) were analysed prospectively and compared with data from the 1995 study. The proportion of patients who received thrombolytic treatment increased from 24% in 1995 to 37% (p = 0.02). All patients received thrombolytics when indicated. The proportion of patients who died in hospital decreased from 25% in 1995 to 16% (p = 0.06). The proportion of patients who were treated within 60 minutes after arrival at the hospital increased from 20% to 67% (p < 0.001). For patients with typical ECG changes at arrival the mean door-to-needle time was 37 minutes. Mean delay from onset of symptoms to treatment was 4.5 hours. The results indicate that improved routines may have increased the proportion of patients receiving thrombolytic treatment and reduced the in-hospital delay. It is possible that a further reduction of delay may be achieved by reducing the pre-hospital delay, or by starting thrombolytic treatment before arrival to hospital.