In our recent paper (Beck et al, 1969) we describe two patients with mild haemophilia who developed inhibitors of factor VIII during replacement therapy to cover major surgery. In both patients the inhibitory activity rapidly waned and they eventually regained their previously mild haemophilic state. Since 1966, Case I has received cryoprecipitate on eight separate occasions for various traumatic lesions. When tested he has had the normally expected rise of plasma factor VIII and his basal level remained at 8%. In July 1969, however, and since our paper went to press, h s patient was treated for a traumatic haematoma in the forearm muscles. He received two doses of 8 packs of cryoprecipitate, each dose containing approximately 800 ml plasma equivalents (units) of factor VIII. The response to the first dose, given at night, was not assayed but after the second dose his factor-VIII level rose from 8% to only 17%. During the next few days it fell to less than 1% and a weak inhibitor, 0.9 units per ml (Biggs & Bidwell, 195g), appeared in his plasma. The haematoma resolved but at a slower rate than would be expected in a normally responsive patient. Six weeks later his factor VIII was 1% but the inhibitor was no longer detectable. Our experience indicates that an inhibitor developing in a mild haemophiliac can recur with subsequent treatment: a situation analogous to that seen in severely affected patients. It is not clear why the inhibitor should have recurred following the last course of treatment and not on previous occasions.Much still remains to be learned concerning possible antigenic variants of factor VIII and the immunological mechanisms of inhibitor development.