The purpose was to evaluate the 10-year mortality in patients with acute chest pain suspected of myocardial infarction with (AMI) and without (non-AMI) confirmed diagnosis and to determine risk factors from the medical history and the diagnosis at discharge. One-thousand eight-hundred and ninety-seven non-AMI patients and 1,401 patients with AMI consecutively admitted to 1 of 16 coronary care units participating in The Danish Verapamil Infarction Study were included. During follow-up, 630 deaths occurred among the non-AMI patients and 415 of these could be classified as cardiac deaths. Multivariate analysis identified the following risk factors containing independent prognostic information about mortality for non-AMI patients: age, previous AMI, sex, and diabetes. In patients with AMI the risk factors were: age, previous AMI, clinical heart failure, diabetes, and angina pectoris. By including the diagnosis at discharge for non-AMI patients in the Cox analysis, the prognostic significance was compared to the variables from the medical history. Only the diagnoses bronchopneumonia, musculoskeletal disorders and observatio sine indicatione therapiae added independent prognostic information. We conclude that non-AMI patients are at high risk for mortal events in the long term. High-risk patients can be identified from their medical history, whereas the diagnosis at discharge only adds limited prognostic information. All non-AMI patients should be carefully evaluated regarding coronary artery disease at the time of discharge in order to improve the risk stratification, treatment and prognosis.