OBJECTIVE To determine the rates of hospital admission for respiratory syncytial virus (RSV) infection among children born at different gestational ages. To assess the theoretical impact of palivizumab prophylaxis on admissions for RSV infection. METHODS Retrospective cohort study of children born in 1991-2000. METHODS Tertiary care university hospital. METHODS Data on all children born during the 10 year period were combined with information on laboratory confirmed RSV infections in these children until the end of 2002. The theoretical impact of palivizumab on RSV associated admissions was estimated by applying the current recommendations for prophylaxis to the study population and using the observed rates of admission in the calculations. METHODS None. METHODS Rates of RSV infection and hospital admission in different subgroups of children. RESULTS Children with chronic lung disease (CLD) were admitted for RSV infection at a rate of 12.0%. The corresponding rates in children born at </=28 or 29-32 weeks gestation were 7.1% and 6.8% respectively. Children born at </=32 weeks gestation accounted for 6.6% of all admissions due to RSV. Of 586 children who would have met the criteria for palivizumab prophylaxis, 27 (4.6%) were admitted with RSV during the presumed prophylactic period. The number needed to treat to prevent one admission for RSV infection was 15 for children with CLD (with a total cost of 75 000) and 43 for children without CLD born at </=32 weeks gestation (with a total cost of 215 000). CONCLUSIONS The rates of hospital admission for RSV infection in premature infants were substantially lower than those in most previous reports from other countries. Determination of the local rates of RSV admissions in different groups of children would be useful in making decisions about the use of palivizumab.