All patients with suspected myocardial infarction admitted to hospital in four Birmingham health districts were studied to test the hypothesis that Asian patients would be overrepresented and Caribbean patients underrepresented compared with the indigenous population. One thousand four hundred and ninety six patients had a final diagnosis of myocardial infarction or severe angina pectoris. The relative risk of admission for Asian men compared with white men aged 45-64 years was 2.65 (95% confidence interval 2.20 to 3.19) and the risk for Asian men was high for both myocardial infarction and ischaemia when analysed separately. The relative risk of admission for Caribbean men compared with white men was 0.53 (95% CI 0.33 to 1.20). The relative risk for Asian women compared with white women in the same age group was 2.58 (95% CI 1.68 to 3.96), but this was due to an excess of admissions diagnosed as ischaemia rather than infarction in the Asian women. For Caribbean and white women the risk of admission was the same, although significantly fewer Caribbean women were admitted with myocardial infarction. The study was undertaken in 1986-87 and population data had to be derived from the 1981 census. The resident population changed in those five years and so the results were recalculated making allowances for these changes in the health districts involved. Based on these data the admission rate for Asian men with suspected myocardial infarction aged 45-64 was nearly twice that for white men (1.8): the relative risk of admission for Asian men compared with white men was 2.04 (95% CI 1.53 to 2.18). For Caribbean men the relative risk compared with white men was 0.45 (95% CI 0.29 to 0.71). For Asian women the relative risk of admission calculated from the adjusted census data resemble that in white women aged 45-64 years. The relative risk for admission with coronary heart disease in Asians is higher than expected work; one possible explanation for this is that the Asian population resident in the area under study was larger than estimated. The single major difference in risk factors was the high prevalence of diabetes mellitus in Asians (19.5% compared with 8.3% for white residents) but this did not wholly account for the excess of admissions from the Asian community.