Case histories are presented of two children in whom thrombocytopenic purpura followed rubella, and a third case is described in which purpura may have been secondary to a modified rubella infection. The serum of two of the children appeared to contain platelet agglutinins. The clinical features of these 2 and 17 other published cases of post-rubella purpura are summarized and discussed. Some clinical and pathologic aspects of acute thrombocytopenic purpura in childhood are discussed, and it is suggested that infections mild enough to pass unrecognized, such as rubella, may precipitate the disease more frequently than is suspected.