A controlled prospective study, in an animal model chosen to simulate portable neonatal radiography, was performed to compare the response of the Philips Computed Radiography (CR) system and conventional 200 speed film-screen (FS) to variation in imaging kVp. Acceptable images were obtained on the CR system over a very wide kVp range. In contrast the FS system produced acceptable images over a narrow kVp range. This ability suggests that the CR system should eliminate the need for repeat examinations in cases where a suboptimal kVp setting would have resulted in an unacceptable FS image. CR technology should therefore be ideally suited to portable radiography especially in situations where selection of correct exposure factors is difficult as in the neonatal nursery.