OBJECTIVE: To compare long term outcome after first myocardial infarction among British patients originating from the Indian subcontinent and from Europe. DESIGN: Matched pairs study. SETTING: Coronary care unit in central Leicester. SUBJECTS: 238 pairs of patients admitted during 1987-93 matched for age (within 2 years), sex, date of admission (within 3 months), type of infarction (Q/non-Q), and site of infarction. MAIN OUTCOME MEASURES: Incidence of angina, reinfarction, or death during follow up of 1-7 years. RESULTS: Patients of Indian subcontinent origin had a higher prevalence of diabetes (35% v 9% in patients of European origin, P < 0.001), lower prevalence of smoking (39% v 63%, P < 0.001), longer median delay from symptom onset to admission (5 hours v 3 hours, P < 0.01), and lower use of thrombolysis (50% v 66%, P < 0.001). During long term follow up (median 39 months), mortality was higher in patients of Indian subcontinent origin (unadjusted hazard ratio = 2.1, 95% confidence interval 1.3 to 3.4, P = 0.002). After adjustment for smoking, history of diabetes, and thrombolysis the estimated hazard ratio fell slightly to 2.0 (1.1 to 3.6, P = 0.02). Patients of Indian subcontinent origin had almost twice the incidence of angina (54% v 29%; P < 0.001) and almost three times the risk of reinfarction during follow up (34% v 12.5% at 3 years, P < 0.001). The unadjusted hazard ratio for reinfarction in patients of Indian subcontinent origin was 2.8 (1.8 to 4.4, P < 0.001). Adjustment for smoking, history of diabetes, and thrombolysis made little difference to the hazard ratio. Coronary angiography was performed with similar frequency in the two groups; triple vessel disease was the commonest finding in patients of Indian subcontinent origin and single vessel disease the commonest in Europeans (P < 0.001). CONCLUSIONS: Patients of Indian subcontinent origin are at substantially higher risk of mortality and of further coronary events than Europeans after first myocardial infarction. This is probably due to their higher prevalence of diffuse coronary atheroma. Their need for investigation with a view to coronary revascularisation is therefore greater. History of diabetes is an inadequate surrogate for ethnic origin as a prognostic indicator.