To evaluate an internal-external quality control program of four automated counters. Every one or two weeks during 14 months, six direct cell parameters were measured in three fresh blood samples in four Coulter counters. The median per parameter of the working day was used to detect inaccuracies and if the participants' internal control program confirmed it, a recalibration of the parameter was performed. In 21 of 22 instances, the internal program confirmed an inaccuracy and a recalibration was done (4 leukocyte and 5 erythrocyte counts, 5 hemoglobins, 7 red-cell volumes). In these four parameters there were no large differences between the lowest and highest counter upon analyzing all results whereas all counters differed from one another in the parameters that cannot be recalibrated by the user (platelet volume, red-cell distribution width). 1. The program contributed to good accuracy and precision within-counters and good concordancy between-counters in the parameters that can be recalibrated. 2. The counter differences in red-cell distribution width were sufficiently large (up to 9%) to affect clinical interpretation. This poses the need of width distribution reference ranges for each counter.