Development of platelet clumps, fusion (VM) and fibrin was studied when an aliquot of platelet-rich plasma (PRP), prepared in siliconed glassware without anticoagulant, was transferred to a glass or siliconed tube and shaken at 37° C. When PRP was not chilled prior to shaking at 37° C in a siliconed tube, clots rarely formed for 12 minutes. Platelet clumps either developed about 1 minute before the clot or did not form at all. If PRP was chilled before shaking, some samples developed platelet clumps within a minute, long before clotting occurred. Either the platelets in these clumps did not fuse, and the clumps disappeared as the sample became warm, or fusion occurred and the clumps persisted. The distinction between clumps caused by chilling and those produced just prior to coagulation, presumably by thrombin, was even clearer in PRP from patients with prolonged clotting times caused by AHF, PTC and PTA deficiency than it was in normal PRP. Chilled samples usually developed early clumps even though clotting was delayed. In unchilled samples, clumping was delayed as well as clotting; this was attributed to sluggish thrombin evolution. When EDTA1 PRP was chilled for several hours and shaken at 37° C in the presence of a physiologic excess of magnesium, clumps formed and the platelets often underwent fusion or VM. No relationship could be found between this type of clumping and the coagulation mechanism, although mercuric chloride and benadryl inhibited both VM produced by cold and that caused by clotting. * Aided by a grant from the American Heart Association and in part by the U.S. Army Medical Research and Development Command, Dept. of the Army, under contract No. DA-49-007-MD-1013.