OBJECTIVE To evaluate the performance of the Coulter MaxM autoanalyser, based on volume, cell conductivity and laser-beam dispersion, and to assess the automated leucocyte differential count (LDC). METHODS A total number of 2,016 blood samples drawn on tri-K EDTA as anticoagulant were studied. The following data were analysed: 1-Intra- and inter-assay inaccuracy, the findings being expressed as mean and variation coefficient (CV%). 2- Correlation with the results attained with other systems (Coulter STKS and Technicon H1); The Pearson's correlation coefficient was used for this analysis. 3- Assessment of the morphologic suspect warning flags, optic microscope examination being used as reference. 4- Stability of the samples along time; the samples were analysed at 2, 6, 24 and 48 hours after withdrawal, and the Scheffe's test was used for comparison of results. 5- Performance speed. RESULTS 1- The intra-assay inaccuracy was found acceptable for all the values except the basophil count (CV: 30.8%). The lack of reproducibility for basophil count is similar to that observed in the Coulter STKS system. 2- Correlation: Optimal LDC was found for neutrophils, eosinophils and lymphocytes. Excellent correlation with the Coulter STKS was seen except for haemoglobin, haematocrit and basophil count. The correlation with the Technicon H1 was better then with the STKS, except for basophil and monocyte counts. The correlation with basophil count was low with every system. 3- Warning flags appeared in 13.4% of the samples, the worse results corresponded to basophils, monocytes, stabs and atypical lymphocytes. 4- Samples stored at 4 degrees C for 24 hr had not significant variations in LDC. Those ones stored at 20-25 degrees C only showed significant variations after 24 hr in the monocyte count. 5- The working speed was 60 samples processed and printed in one hour. CONCLUSIONS The Coulter MaxM autoanalyser is a useful machine for laboratories and provides a 5-class LDC plus a sound screen for leucocyte morphologic anomalies.