The aim of this paper is to compare the image quality of different CT units at four centres using first the same dose per slice, and then the dose per slice used when performing a standard abdominal acquisition. The results show that the detection of a 7 to 9 mm supra-slice object (i.e. larger than the slice thickness) of 0.5% can be obtained at a comparable dose level. The main difference between the performances of the CT appears when trying to detect a contrast level of 0.3%. When comparing the CT using the clinical acquisition parameters chosen, the same conclusions can be drawn in spite of the fact that the CTDI values varied by 65%. The most important result is the comparison of the effective dose which shows differences up to almost 100% between the centres involved. This paper confirms that the DLP concept is a powerful tool for the optimisation of CT examination protocols.