The long-term outcome of myocardial infarction (MI) with bundle branch block (BBB) was studied retrospectively by a direct questionnaire and a registry enquiry in order to define the prognostic significance of data obtained during hospitalisation and to discuss the possible indications of permanent pacing in these patients. Out of 2 720 acute MI hospitalised between October 1969 and April 1977, 231 and BBB (unknown before infarction): 58 right, 53 left and 120 bilateral BBB. 113 patients survived the acute phase of MI and 111 patients were followed up for 72 +/- 24 months: 80 patients died, and 30 survive - a global survival rate of only 13% at 6 years. The post-admission mortality rate was not related to the type of BBB or the site of infarction. It was significantly higher in patients with previous myocardial infarction and in a sub group of patients without advanced AV block in the acute phase who had severe cardiac failure (Classes III and IV, Killip). The hospital mortality was higher in patients with advanced AV block in the acute stage (62,5% compared to 45,9%) in patients without advanced AV block, p < 0,025. On the other hand, the post-admission mortality was not significantly different in these two sub groups (77,8% compared to 69,8% : NS). Stokes-Adams syncope was rarely authentified in the post-admission course of the patients. An increased risk of secondary sudden death directly related to a conduction defect has not been proved. The indications for permanent pacing therefore remain uncertain. They should not be widened further than the indications for permanent pacing in chronic degenerative block.