Essential thrombocythaemia (ET) is a disease associated with an elevated risk of thrombosis. This study evaluated the efficacy and safety of pipobroman (PB) in the long-term control of ET patients who had, at diagnosis, one or more of the following currently known risk factors for thrombosis or haemorrhage (high-risk patients): age > 60 years, history of thrombosis or haemorrhage, platelets >1000 x 10(9)/l. From 1978 to 2000, with a median follow-up of 10 years, 118 previously untreated high-risk ET patients (median age 62 years, range 25-82), were treated with PB at the starting dose of 0.8-1 mg/kg/d. All patients reached a platelet count <600 x 10(9)/l and 91% achieved a platelet count <400 x 10(9)/l. During follow-up, 13 patients had thrombosis, with a 10-year cumulative risk of 14%. Acute myeloid leukaemia, myelofibrosis and solid tumours occurred in three, two and seven patients with a 10-year cumulative risk of 3%, 2% and 7% respectively. Actuarial survival at 20 years was 64% and the standardized mortality ratio was 1.1 (95% CI: 0.7-1.7), not statistically different from the general population (P = 0.54). Age was associated with a higher risk of death (P = 0.00009) and thrombosis (P = 0.003). The duration of PB treatment did not correlate with the occurrence of second malignancies. This study, with a median follow-up of 10 years, demonstrates that pipobroman is effective and well tolerated. The low cumulative 10-year risk of thrombosis, leukaemia and solid tumours indicates that pipobroman is an adequate treatment for patients with high risk ET.