This study was conducted to determine whether the doses for the radiographic examination of the lateral lumbar spine changed as a result of the introduction of a hospital-wide picture archiving and communication system (PACS). Doses were measured by thermoluminescent dosimeters (TLD) and dose-area product (DAP) meter readings for 100 patient examinations using a 300-speed conventional film/screen system and for 96 patient examinations when PACS was fully operational. Radiographic technique, exposure factors and patient characteristics were noted and effective doses were calculated, and a comparison was made of all variables. No significant differences between conventional and PACS working were found in surface entry and effective doses for single views of the lateral lumbar spine, but there was a 20% reduction in DAP readings with PACS. However, when summed doses for all images, including rejects, required to demonstrate the lateral lumbar spine for each patient were compared, PACS was found to be associated with significantly lower surface entry (TLD) dose, DAP reading and effective dose (28%, 36% and 16%, respectively) than conventional film. For single images of L1-5, when PACS was in use, there was a significant reduction in the DAP readings and increases in the area of the film/plate irradiated, the focus-to-skin distance and the focus-to-film distance. In addition, significantly fewer lumbosacral junction views were undertaken when PACS was in use. Since many confounding factors may have influenced the results over the period of dose measurement, regression models were used to determine the significance of PACS. These models showed that the use of PACS was not significant in causing any differences in the dose for single images as compared with when film was used, but was significant in the resulting total dose reductions for the examinations.