The radiation dose-limiting toxicity from radioimmunotherapy has been myelotoxicity in the absence of bone marrow reconstitution (transplantation). Myelotoxicity can be assessed directly by biopsy examination of the bone marrow and indirectly by peripheral blood counts. In patients with B-cell malignancies, thrombocytopenia has been the initial and most severe manifestation of 131I-Lym-1 radiation toxicity from treatment. Manifestations of myelotoxicity varied greatly among the patients and from one treatment dose to another in the same patient, suggesting that additional factors were present. There was an increased likelihood of Grade 3-4 hematopoietic toxicity after 131I-Lym-1 treatment if the patient had peripheral blood cell abnormalities before undergoing 131I-Lym-1 treatment. Fractionation of the total 131I-Lym-1 dose was associated with less toxicity. In many patients, myelotoxicity could not be explained by marrow radiation dose (0.36 +/- 0.13 rads per administered mCi) from 131I-Lym-1 in the blood and body alone. Bone marrow examination and 131I-Lym-1 imaging usually provided evidence for additional marrow radiation from 131I-Lym-1-targeting of marrow malignancy and also for residual toxic effects from prior treatment in these patients. Immunohistologic and imaging examination of the bone marrow performed with the intended treatment antibody allowed assessment of extent of marrow malignancy and prediction of degree of myelotoxicity from subsequent treatment. Treatment programs (and protocols) for radioimmunotherapy should incorporate these methods into the decision process. Larger amounts of 131I-Lym-1 can be used in patients selected to have relatively normal peripheral blood cell counts and normocellular bone marrows uninvolved by the malignancy. These observations appear to be relevant to the maximum tolerated dose in radioimmunotherapy for other malignancies as well.