Concern over reported large radiation doses leading to a high cancer risk for paediatric CT patients has prompted considerable investigation in paediatric CT. The recent release of software from Germany has allowed effective doses to be calculated from CT protocol information and radiation measurement for standard paediatric patient sizes for both sexes. An initial study has been undertaken in nine radiology departments, four of which were dedicated paediatric departments, for routine chest and abdominal CT procedures. The dose calculation software is based on Monte Carlo simulation of X-ray conditions during a CT procedure and utilized a 'tomographic' phantom model of a 7-year-old child and an 8-week-old baby to allow calculation of organ dose and hence effective dose. Results of the survey indicate that effective doses were higher for females than males, and higher for abdominal procedures. Slightly higher effective doses were calculated for the child compared to the baby. All centres but one recorded lower effective doses with their current protocols than if they had used recommended CT protocols found in the literature. Analysis of the survey data indicates that scan parameters are the main cause of dose variations, although the type of scanner can affect dose by a factor of 2 (when comparing different units) as well as variation in anatomy scanned in protocols. Dose reduction appears to be most closely linked with reduced mAs and increased pitch as expected. The calculation of effective dose appears to be a key factor in assessing CT protocols, particularly for paediatric patients.