OBJECTIVE: To determine whether the separation rate from the hospital for children aged 0-15 years in Victoria was higher for those resident in the country area of the State in comparison with the metropolitan area and, if it was, to investigate possible explanations. DESIGN: Discharge data from all public hospitals in Victoria for children aged 0-15 years for the financial years 1988-89, 1989-90 and 1990-91 were analysed with detailed analysis being done on the 1990-91 data set. Discharge rates were determined according to the local authority area of residence. Patients were grouped according to Diagnosis Related Groups (DRGs) version 5. RESULTS: Children living in the country area showed a separation rate of 50 per cent greater than that for the metropolitan area. Separation rates for local authority areas were remarkably constant over the three years. Country local authority areas with the highest separation rates had separation rates for asthma and bronchitis (DRG 98), almost four times that of metropolitan residents, and for otitis media and upper respiratory infection (URI), a rate almost ten times that of metropolitan residents. CONCLUSION: It is suggested that variation in medical practice was the most likely explanation for the observed differences.