We performed an open, nonrandomized, multicenter phase-II trial to evaluate the efficacy and toxicity of 1 year of treatment with the oral iron chelator deferiprone in 38 mainly nonthalassemic patients with transfusional iron overload. Initial serum ferritin varied between 996 and 11.644 micrograms/l. Patients were treated with 3-6 g of deferiprone daily. Mean urinary iron excretion (UIE) in 36 evaluable patients was 21.0 mg/24 h and was significantly higher in the patients with thalassemia than in those with myelodysplasia. Negative iron balance was achieved in 20 patients (56%). The median duration of treatment was 10 months; due to side effects and other causes only 20 patients completed 1 year of treatment. Mean serum ferritin levels decreased from 3563 micrograms/l at the start of the trial to 2767 micrograms/l at 6 months (26 patients, p < 0.004) and to 2186 micrograms/l at 12 months (20 patients, p < 0.005). Serum ferritin levels normalized in two patients who were no longer transfusion dependent. Deferiprone was clearly not effective in three patients (two with myelofibrosis, one with myelodysplasia). One patient with myelodysplasia developed agranulocytosis after 12 months of treatment; this was rapidly reversible after stopping deferiprone. Three patients had a mild and transient decrease in white blood cell count. Other side effects leading to withdrawal from the trial consisted mainly of nausea (3 patients), arthralgia (2), and skin rash (1). No clinical signs of zinc deficiency were seen, although zinc excretion was increased in three patients. No changes were seen in liver enzymes, creatinine, antinuclear factor, T-cell subsets, cardiac function, visual acuity, and audiogram. Although our results confirm deferiprone as an effective iron chelator in patients with thalassemia and in some patients with other forms of iron overload, there is still some concern about the safety of this drug, which therefore, at this time, should be used exclusively in well-controlled clinical trials.