BACKGROUND: Emergency admission rates have been rising rapidly in Britain. Studies defining the underlying factors are needed. AIM: To determine the principal diagnoses, demographic, and socioeconomic factors associated with emergency medical admissions. METHOD: Cohort study based on the Greater Glasgow Health Board population of 810,423 adults. A fully anonymized dataset linkage of 43,247 adult emergency admissions to Glasgow medical beds in 1997 was obtained. Emergency admission rates were analysed by diagnosis, age, sex, Carstairs' deprivation category, and by individual general practices (after adjustment for other factors). RESULTS: The commonest principal diagnoses were chest pain (9.6%), chronic obstructive airways disease (5.6%), angina (5.4%), heart failure (4.1%), and acute myocardial infarction (3.9%). Twenty-one per cent of patients were coded as having 'ill-defined signs or symptoms'. Emergency medical admission rates rose with the age of the patient, doubling with every two decades' age increase. Admission rates for patients from deprived areas were twice those from affluent areas. Males were more frequently admitted than females (adjusted odds ratio = 1.19). After adjustment for age, sex, and deprivation, the general practices' emergency medical admission rates showed an almost twofold difference between the top and bottom deciles. CONCLUSION: Emergency medical admission rates are higher among the elderly, males, and deprived populations. This has implications for equitable resource distribution in the National Health Service. Admissions for exclusion of myocardial disease were common; however, myocardial infarction was not the final diagnosis in two-thirds of these patients. The large variation between the general practices' admission rates requires further investigation.