Using autologous 111In-labeled platelets, platelet kinetics and the sites of platelet destruction were assessed in 16 normal subjects (13 with and three without spleens), in 17 studies of patients with primary autoimmune thrombocytopenic purpura (AITP), in six studies of patients with secondary AITP, in ten studies of patients with AITP following splenectomy, and in five thrombocytopenic patients with myelodysplastic syndromes. In normal subjects, the spleen accounted for 24 +/- 4% of platelet destruction and the liver for 15 +/- 2%. Untreated patients with primary AITP had increased splenic destruction (40 +/- 14%, p less than 0.001) but not hepatic destruction (13 +/- 5%). Compared with untreated patients, prednisone treated patients did not have significantly different spleen and liver platelet sequestration. Patients with secondary AITP had similar platelet counts, platelet survivals, and increases in splenic destruction of platelets as did patients with primary AITP. In contrast, patients with myelodysplastic syndromes had a normal pattern of platelet destruction. In AITP patients following splenectomy, the five nonresponders all had a marked increase (greater than 45%) in liver destruction compared to five responders (all less than 40%). Among all patients with primary or secondary AITP, there was an inverse relationship between the percent of platelets destroyed in the liver plus spleen and both the platelet count (r = 0.75, p less than 0.001) and the platelet survival (r = 0.86, p less than 0.001). In a stepwise multiple linear regression analysis, total liver plus spleen platelet destruction, the platelet survival and the platelet turnover were all significant independent predictors of the platelet count. Thus platelet destruction is shifted to the spleen in primary and secondary AITP. Failure of splenectomy is associated with a marked elevation in liver destruction. The magnitude of spleen and liver destruction appears to be of considerable importance in the severity of the disease, as reflected in the platelet survival and platelet count.