In this retrospective study, we evaluated the clinical features and the effects of various treatment modalities on the clinical course in patients diagnosed with idiopathic thrombocytopenic purpura (ITP) at our center between 1984-2000. We retrospectively examined the medical records of 321 (229 females, 92 males) ITP patients. One hundred and seventy-one (53.3%) patients were lost to follow-up. When evaluating the clinical features, all 321 patients were included; however, when the response to treatment modalities was evaluated only 150 patients followed up regularly were considered. The median age of the patients on initial diagnosis was 34 years (range: 14-78). At initial diagnosis, 235 (73.2%) patients had signs of bleeding. Of patients diagnosed with ITP initially, six later turned out to have systemic lupus erythematosus (SLE) and two myelodysplastic syndrome (MDS). The median follow-up of 150 patients followed up regularly was 30 months (range: 4-396). One hundred and thirty-seven of these subjects had an indication for treatment and 94.2% of them were administered either standard or high-dose steroids as the first-line therapy. Complete remission (CR) was defined as any platelet count >100,000/mm(3) lasting for 3 months or longer without treatment. CR was achieved in 51.9% of the patients given steroids as the initial therapy. During a median follow-up of 33 months, relapse occurred in 58.2% of these patients, and after a median follow-up of 11 months the rest of them were still in remission. Ninety-eight patients followed up regularly were administered second-line therapies. CR was obtained in 44.4% of the patients who used steroids as second-line therapy. Within a median follow-up of 15 months, 20.8% of these patients relapsed. Splenectomy was performed in 76 patients and CR was obtained in 68.4% of the regularly followed up patients. Relapse occurred within a median of 96 months in 15.4% of the patients who had CR. Kaplan-Meier curves showed that the duration of CR obtained by splenectomy was significantly higher than that obtained by steroids (p<0.001). The 10-year disease-free survivals in patients who used steroids and who underwent splenectomy were, respectively, 13% and 58%. In our adult ITP patients, steroids induced nearly similar rates of CR both as first-and second-line therapies. Splenectomy seems to be effective in patients unresponsive to steroids. The duration of CR obtained by splenectomy is significantly longer when compared with the duration of CR obtained by steroid therapy.