The results of splenectomy in 33 patients studied in series 1-7 (Table I) are shown in Table II. It seems encouraging that only one of 18 patients showing ‘splenic’platelet destruction did poorly. On the other hand, platelets were restored to normal by operation in five of eight patients in whom platelet destruction was thought to occur primarily in the liver. Results of 206 splenectomies in Najean’s series are given in Table III. In agreement with the experience of groups 1-7, only 6% of patients with ‘splenic’platelet destruction did poorly. By contrast, however, eight of 11 patients in whom platelet destruction was ‘hepatic’received little or no benefit from operation. It may be significant that eight of 15 patients in series 1-7 with ‘hepatic’platelet destruction were splenectomized but only 11 of 75 such cases were operated in the series of Najean & Ardaillou (1971). The latter patients were, therefore, more highly selected. It is only a modest oversimplification to state that, in the opinion of most workers, clear-cut splenic sequestrations of platelets favours a successful outcome of splenectomy. Only Najean’s group, however, goes so far as to advise against operations in patients with ‘hepatic’destruction of platelets. One may question whether, in the latter series, a sdcient number of such patients were splenectomized to fully justlfy this conclusion.