As the disappearance of platelets in extracorporeal circulation has been established as a problem in surgical and experimental applications, an investigation was undertaken to determine the fragility of normal human blood platelets. A study of recognized procedures for platelet anticoagulation and storage dictated handling procedures and preparation of a platelet-rich plasma (PRP). An attempt was made to eliminate variables of platelet function—aggregation, adhesion, etc.—by careful choice of donors and adherence to recommended handling techniques. Viscometry equipment, originally used for shear of whole blood, was modified for application to PRP. Analyses were developed for determination of platelet destruction as a result of shear stress. Release of the platelet enzyme acid phosphatase and Coulter Counter platelet enumeration provided consistent, if not exactly comparable, destruction data. Phase microscopy platelet enumeration was not successful in determining platelet destruction. Results of applied shear to the PRP indicated a critical threshold shearing stress for platelet destruction in the range of 15-2 dynes/cm. This threshold is approximately one-tenth of that found for red blood cells through previous experimentation in this laboratory. Through varying of viscometer configuration, it was determined that the platelet shear destruction was not surface dependent. It was, however, increased by turbulent flow. A secondary set of results was obtained when a "storage lesion" developed prior to shear in ineffectively anticoagulated PRP. With appearance of this lesion, platelet destruction was drastically increased and inconsistent. The storage lesion correlated with increased acid phosphatase level in the platelet-poor plasma controls and with platelet aggregation observed after shear.