The Clinically Applied Analytical Cytometry (CAAC) project was a concerted action project in a working group measuring DNA in breast cancer. An internal quality assurance programme was established on a voluntary basis to determine the level of concordance of DNA measurements between participating laboratories. Three rounds were achieved within the time scale of the project. For each round three slides (2 prepared with Feulgen, one unstained for "in house" preparation) bearing a population of human liver cells were sent to participating laboratories. The institutions were asked to measure 200 diploid, 100 tetraploid and 50 octaploid cells by means of the image cytometry system present in the laboratory. In the third round tumor cells were added. The features integrated optical density (IOD) and AREA were reported. In the three rounds the number of participating laboratories was 11, 14 and 11, respectively. The interlaboratory variation expressed as the CV of IOD for the three rounds ranged from 2-18%. Calculation of the 4c/2c and 8c/2c ratios revealed a high precision for most of the instruments. Comparison of measurements in specimens stained in the participating and central laboratory showed similar CV values. Measuring 200 nuclei in image analysis is too low a number to obtain a reliable estimate of the S-phase fraction. In conclusion, the interlaboratory intercomparison of DNA measurements performed on different instruments is well feasible and could facilitate improvement in quality standards.