The very-long-term follow-up of patients initially included in the PVSG protocols provides useful information. The excess risk of cancer after chlorambucil appears to persist for 5 years after stopping this treatment. The risk of leukaemia induced by marrow suppression (32P or chemotherapy) was marked before the 10th year, but low thereafter. Phlebotomy is unacceptable as permanent treatment because of the poor clinical tolerance and the frequency of vascular complications. This treatment is also associated with a risk of early progression towards myelofibrosis with myeloid splenomegaly. In the very long term, 15 years or more after the diagnosis, this complication is the major clinical risk, affecting almost 50% of our patients surviving at this time. The prevention of this type of complication could constitute one of the objectives of future protocols dealing with this disease.