Platelet counts in platelet-rich plasma without hematocrit dependent correction were performed by following rapid and simple steps: 1. pre-dilution of 20 microliter of whole blood by an isotonic solution 1:25; 2. stabilized low-speed centrifugation with 55 g for 5 minutes; 3. final dilution 1 : 5000; 4. enumeration by use of a TOA platelet counter PL-100 which has been technically improved in comparison to similar machines. Erroneously high results were obtained after a too short or too low centrifugation. As reason for this artifical small pulses due to disturbances of the flow patterns around the aperture (so-called vortex-effect) can be assumed having been caused by large-volumed erythrocytes and leukocytes in the suspension. The routinely used procedure was reliable for all platelet ranges, especially in thrombocytopenias between 100 X 10(9)/l and 25 X 10(9)l. In lower ranges comparisons with visual counts are essential.