An understanding of the natural history of childhood chronic idiopathic thrombocytopenia purpura (ITP) could contribute to a rational therapeutic approach to its treatment, which remains controversial. In our retrospective study of 92 children with ITP, 22 had a chronic course and were followed for 3-14 years (median 8.6 years). Treatment, when indicated, was individualized: 4 patients (18.2%) did not receive any treatment, 14 (63.6%) received steroids only, while 4 (18.2%) were treated with steroids and one of the following: high-dose gamma globulin (4 patients), splenectomy (2 patients) or immunosuppressive therapy (2 patients). During follow-up, 14 patients (63.6%) achieved complete remission, 5 (22.7%) partial remission and only 3 (13.5%) remained severely thrombocytopenic, with minimal bleeding tendency. Eleven patients (50%) responded to the initial prednisone course (1-5 mg/kg/day), but showed a marked decrease in platelet count when steroids were tapered off. In view of the high rates of complete and partial remission and the mild course of the few non-responding patients, it is suggested that with adequate supportive therapy, follow-up problems and fatalities can be kept to a minimum. We believe that aggressive therapy, such as splenectomy, should be reserved for the rare symptomatic and severely thrombocytopenic patient.