Aspirin, intravenous nitrates and fibrinolysis were being used by 1986 in Girona, Spain. These combined factors should be reflected in myocardial infarction patients' outcome. We assessed changes in 28-day and 3-year survival after a first myocardial infarction between 1978-85 and 1986-88 in the REGICOR (Registre GIroní del COR) registry. This included 1216 consecutive patients with a first transmural myocardial infarction (834 in 1978-85 and 372 in 1986-88). Their 28-day and 3-year mortality rates were 14.6% and 8.8% respectively. Although patients admitted in the second period were more frequently hypertensive and diabetic, a history of angina was less common in patients admitted between 1978 and 1985. After adjusting for diabetes, hypertension, age, and sex, the relative risk of 28-day mortality of those admitted in the second period was 0.65 (95% confidence interval 0.42-0.99). The lower severity, as measured by Killip class of patients in the second period, was the main (confounding) variable responsible for this protective effect. Three-year mortality of those surviving 28 days in 1978-85 (8.3%) did not differ from 1986-88 (8.3%). In the second study period hospitalized patients with myocardial infarction in Girona, Spain showed a better 28-day survival. It is possible that therapeutic and diagnostic refinements, together with other factors not controlled in the present study, have resulted in such an improvement. However, 3-year mortality remained unmodified among those surviving 28 days.