Quantitative aspects of platelet agglutination and inhibition of clot retraction by the antibody of quinidine purpura were described. The reactions appeared to depend on formation of types of antibody-quinidine-platelet complexes which could fix complement but complement was not necessary for these reactions. Complement fixation was at least 10 times more sensitive than platelet agglutination or inhibition of clot retraction for measurement and detection of antibody activity. Although it has been considered that antibodies of drug purpura act as platelet lysins in the presence of complement and that direct lysis of platelets accounts for development of thrombocytopenia in drug purpura, the present study suggests that attachment of antibody produces a change in platelets which is manifested in vitro only by increased susceptibility to non-specific factors which can alter the stability of platelets in the absence of antibody. The attachment of antibody to platelets in vivo may only indirectly affect platelet survival. In contrast to human platelets, dog, rabbit, and guinea pig platelets, and normal or trypsin-treated human red cells did not agglutinate, fix complement, or adsorb antibody; and intact human endothelial cells did not fix complement or adsorb antibody. Rhesus monkey platelets were not agglutinated by the antibody but did adsorb antibody and fix complement although their activity in these reactions differed quantitatively from that of human platelets. Cinchonine could be substituted for quinidine in agglutination and inhibition of clot retraction reactions but quinine and cinchonidine could not. Attempts to cause passive anaphylaxis in guinea pigs with the antibody of quinidine purpura were not successful.