A computed tomography radiation dose survey was performed within our enterprise using three age-based paediatric phantoms representing a 1, 5 and 10 years old. Twenty-seven scanners were surveyed with volume computed tomography dose index and dose length product data collected for head, chest and abdomen-pelvis protocols at each age. Reconstruction method e.g. filtered back projection (FBP) or iterative (IR) was also recorded. About two-thirds of the 1 year old FBP chest scans exceeded the national Baby diagnostic reference level (DRL). A small number of scanners also exceeded the national Child DRL for the 1 and 5 years old phantoms. Only about half of the phantom protocols showed a difference of statistical significance between FBP and IR scanners. The results suggested the need for optimisation work at a number of sites. It was determined that the proposed local (i.e. enterprise-wide) DRLs are presented best in terms of weight or girth rather than age.