Two techniques for the preparation of platelet concentrate (PC), the standard platelet-rich plasma (PRP) and buffy coat (BC) methods, were compared in nine paired studies with regard to platelet harvest, white cell (WBC) contamination, and PC quality after 5 days of 22-degrees-C storage. Platelet harvest using the BC method averaged approximately 56 percent of the whole blood level (6.2 x 10(10)/concentrate), which was less than the 76 percent achieved with the PRP-PC method (8.7 x 10(10)/concentrate). An additional 5 units collected into an experimental siphon bag for BC-PC processing showed improved platelet harvest (6.7 x 10(10)/concentrate, or approx. 70% of whole blood). WBCs remaining in the BC-PC averaged 0.19 x 10(8), per unit compared to 3.6 x 10(8) per unit for PRP-PC. Buffy coat processing produced red cell (RBC) units with 50 percent of the WBC contamination of conventionally prepared units (9.8 +/- 6.2 x 10(8)/unit vs. 18.9 +/- 7.1 x 10(8)/unit). The siphon bag further reduced WBC levels in the AS-3 RBC units (6.4 +/- 3.7 x 10(8)/unit). In vitro studies performed on Days 1 and 5 after collection showed no significant differences in platelet metabolic and biologic function or cell integrity. Beta-thromboglobulin and surface glycoprotein levels, indicators of platelet activation and membrane alteration, respectively, did not differ significantly in the PRP-PC and BC-PC; nor was lactate production higher in PRP-PC, despite the substantially higher WBC counts. Autologous in vivo platelet viability determinations were performed by using concurrent transfusion of In-111-labeled freshly drawn platelets and Cr-51-labeled stored platelets. Paired t test analysis of BC-PC versus PRP-PC indicated no significant differences in platelet recovery and survival after 5 days of 22-degrees-C storage in polyolefin containers. Therefore, these studies confirm the equivalence of PC quality, comparable platelet harvest with the siphon bag, and decreased WBC contamination with the BC method.