BACKGROUND: The management of acute myocardial infarction (AMI) has improved over the last 50 years with the more frequent use of effective medicines and procedures. The clinical benefit of the specialty of the attending physician is less clear. The United Kingdom National Service Framework for coronary heart disease (CHD) suggested that patients with CHD are likely to benefit from cardiological supervision. We set out to assess the effect of access to cardiologists on survival among AMI patients admitted in two UK hospitals. METHODS: The study was conducted in a university hospital and a district general hospital in England. Information was obtained on age, sex, ethnicity, Carstairs socioeconomic deprivation category derived from postcode of residence, comorbidity, distance from hospital and medication from all patients admitted with acute myocardial infarction in two acute trusts between July 1999 and June 2000. Record linkage to subsequent Hospital Episode Statistics and Registrar General's death records provided follow up information on procedures and mortality up to eighteen months after admission. Cox proportional hazard models were used to investigate the main hypothesis controlling for confounding. The main outcome measure was 18-month survival after myocardial infarction. RESULTS: Access to a cardiologist was univariately associated with improved survival (hazard ratio 0.16, 95% CI 0.10 to 0.25). This effect remained after controlling for the effect of patient characteristics (hazard ratio 0.22, 95% CI 0.14 to 0.25). The effect disappeared after controlling for access to effective medication (hazard ratio 0.70, 95% CI 0.33 to 1.46). CONCLUSIONS: Access to a cardiologist is associated with better survival compared to no access to a cardiologist among a cohort of patients already admitted with AMI. This effect is mainly due to the more frequent use of effective medicines by the group referred to cardiologists. Hospitals may improve survival by improving access to effective medicines and by coordinating care between cardiologists and general physicians.