Ten patients scheduled to receive intensive chemotherapy were plateletapheresed and the platelet-rich plasma was frozen with 5 percent dimethyl sulfoxide at –80 to –95° C until needed. Paired comparisons of frozen autologous platelets with fresh single-donor platelets were made in seven patients using corrected platelet increments at 1 and 24 hours, and pre- and posttransfusion bleeding times. In vitro tests of 12 units of plateletrich plasma before and after freezing included platelet factor 4 (PF4) secretion, malondialdehyde production, and electron microscopic evaluation of morphology. Fresh platelets provided significantly better 1- and 24-hour corrected increments compared with frozen autologous platelets. In only one case of alloimmunization did frozen autologous platelets provide a better increment than fresh platelets. Bleeding times after transfusion showed no consistent improvement regardless of type of transfusion or platelet count. Secretable PF4 remained constant after freezing, but malondialdehyde production fell significantly. Platelets showed considerable structural damage with 33 percent balloon forms counted after thawing, compared to less than 1 percent before freezing. Except in the case of alloimmunization, frozen autologous platelets are inferior to single-donor fresh platelets, and are significantly damaged in the freezing process.