The use of digitalis in patients who have sustained an acute myocardial infarction has been controversial. A number of reports have suggested that the use of the glucoside actually can increase mortality. In all previous reports, digoxin has been the glucoside in use. In 620 patients admitted to Hamar Hospital from January 1, 1982 to December 31, 1984 with an acute myocardial infarction, 159 were using digitoxin at entry. Their survival until January 1, 1985 was compared with the rest of the population, not using digitoxin. The patients on digitoxin had a significantly higher mortality than those who were not, with a relative risk ratio of 2.33. However, major differences in baseline risk factors existed, and a multivariate Cox regression analysis was employed to adjust for the inequalities in baseline covariates. Only variables available at the entry of the index infarction were used for adjustment. Serum sodium, serum creatinine, age and no smoking were identified in a backward selection procedure as independently influencing mortality. By adjusting for these variables the relative risk ratio was reduced to 1.41, but digitoxin still exerted a significant influence on mortality. It is concluded that the results obtained with the use of digitoxin during a myocardial infarction is similar to previous reports with digoxin. Most of the excess mortality in the digitoxin group can be accounted for by inequalities in baseline characteristics. However, the possibility that a small excess in mortality is due to the digitalis glucoside cannot be excluded.