Over a 24 month period on six paediatric wards of different designs 169 cases of possible hospital-acquired respiratory virus infection were investigated. A variety of viruses was isolated from 82 cases, the most common being respiratory syncytial virus, influenza, parainfluenza, adenoviruses and rhinoviruses. A further 73 children developed respiratory symptoms between 3 and 300 days after administration but viruses were not demonstrable by the techniques used. These children were thought to have hospital-acquired infection nonetheless. Thirteen children were shown not to have acquired infection as the cause of their intercurrent illness. Most acquired infections occurred where toddlers were in cots in open wards. Children with trauma, including non-accidental injury, congenital malformations, mental retardation, failure to thrive or neoplasia were most likely to become infected. Almost 20% of children suffered from croup or lower respiratory tract illness as a result of their acquired infection. The figure was 41% if those less than 12 months old were considered alone. Most episodes settled quickly but in a few children investigations or surgery were delayed for a few days.