The present paper describes a flow cytometric method for assay of platelet aggregates (PAg) in blood. This method combines and simplifies previously reported techniques, simultaneously enumerating PAg formed upon platelet activation, their expression of activation marker CD62P (P-selectin), and their content of bound leukocytes (LPAg). The sensitivity of this method to low levels of agonists (ADP, collagen) is compared to conventional aggregometry and some features of platelet-leukocyte interaction are explored. The results were: (1) ADP or collagen induced a dose-dependent increase in PAg number and corresponding decline in free platelets. The ED50 for ADP (0.15 microM) and for collagen (0.2 microg/mL) was about 1/20 the ED50 found by aggregometry, indicating 20-fold greater sensitivity. (2) At higher concentrations, the fraction of PAg with bound leukocytes (LPAg) increased to 60-70%. This rise correlated with PAg size and CD62P expression, but not with the number of PAg formed. (3) The response of whole blood (WBD) to agonists was qualitatively different from that of platelet-rich plasma (PRP): in WBD the population of CD62P+ PAg was much higher than in PRP and the population of CD62P+ free platelets was much lower. This implies that leukocytes rapidly recruit activated platelets. (4) Desmopressin (DDAVP) at 5 nmol/L induced a significant rise in activated (CD62P+) PAg and platelets, even though no effect of DDAVP could be detected by conventional aggregometry; this further confirms that DDAVP acts directly on platelets. (5) Plasma samples from TTP patients induced a rise in PAg when added to normal PRP, though little or no effect could be detected by aggregometry. In summary, the flow cytometric method described here appears useful for detecting low levels of platelet activation and provides information on platelet leukocyte interaction, potentially important in identifying and differentiating thrombogenic states. Since it is rapid and economical, it is well suited for clinical implementation.