We conducted a nonrandomized prospective phase II study of thalidomide in anemic patients with myelofibrosis with myeloid metaplasia (MMM), with or without preceding polycythemia vera or essential thrombocythemia, with a primary aim to improve anemia. Thalidomide was given in escalating doses with a target dose of 800 mg daily, but the median dose of thalidomide that was actually tolerated was 400 mg daily. Fifteen patients were entered into the study and 14 were evaluable for response. Five of 14 (36%) patients discontinued thalidomide before 3 mo because of side effects, and none of these five patients had a response at the time when thalidomide was stopped. When evaluated after 3 mo of therapy, none of the remaining nine patients exhibited a discernible clinical response. Three patients showed progressive disease defined as > 50% increase in the need for red cell transfusions. Treatment was poorly tolerated, with all patients reporting side effects of thalidomide, the most prominent being fatigue documented in 80% of patients. Two patients died while on study, one from acute myelogenous leukemia and one from pneumonia. We conclude that thalidomide given in doses employed in the treatment of multiple myeloma gives no clinically relevant hematological effects in advanced MMM and is hampered by a very high incidence of side effects.