The Coulter Counter model S-Plus measures the erythrocyte, leukocyte, and platelet counts, and the mean cell volume of erythrocytes and platelets. Two additional features are novel to the instrument: erythrocyte distribution width (RDW), an index of erythrocyte volume heterogeneity, and mean platelet volume. For 677 subjects, we compared these measurements with complete blood counts obtained by a standard instrument, platelet counts from platelet-rich plasma (PRP), and histograms of erythrocyte and platelet volumes generated by free-standing equipment. The two methods correlated well (r = 0.97) for the complete blood count and the mean platelet volume. The whole-blood platelet count correlated well (r = 0.97) with the PRP platelet count, with whole-blood counts on average 5% higher than PRP counts. The whole-blood index of erythrocyte volume heterogeneity, RDW, correlated less well (r = 0.81) with the erythrocyte coefficient of variation measured by histogram. RDW was abnormal for three of 100 normal subjects with normal histograms. RDW was less sensitive (78% vs. 97%) and less specific (94% vs. 97%) than erythrocyte histograms in distinguishing iron deficiency from heterozygous alpha- or beta-thalassemia. We conclude that the Coulter Counter Model S-Plus reliably measures complete blood cell count, platelet count, and mean cell volumes, but analyzes erythrocyte volume distribution less well than does a histogram.