The purpose was to compare the image quality and patient dose between 4- and 16-row MDCT units and to evaluate the dispersion of the dose delivered for common clinical examinations. Four 4- and 16-row MDCT units were used in the study. Image noise levels from images of a CatPhan phantom were compared for all units using a given CTDI(vol) of 15.0+/-1.0 mGy. Standard acquisition protocols from ten centres, shifted from 4- to 16-row MDCT (plus one additional centre for 16-row MDCT), were compared for cerebral angiography and standard chest and abdomen examinations. In addition, the protocols used with 16-row MDCT units for diagnosis of the unstable shoulder and for cardiac examinations were also compared. The introduction of 16-MSCT units did not reduce the performance of the detectors. Concerning the acquisition protocols, a wide range in practice was observed for standard examinations; DLP varied from 800 to 5,120 mGy x cm, 130 to 860 mGy x cm, 410 to 1,790 mGy x cm and 850 to 2,500 mGy x cm for cerebral angiography, standard chest, standard abdomen and heart examinations, respectively. The introduction of 16-row MDCT did not, on average, increase the patient dose for standard chest and abdominal examinations. However, a significant dose increase has been observed for cerebral angiography. There is a wide dispersion in the doses delivered, especially for cardiac imaging.