Myelosuppression is the dose-limiting side effect of most forms of radioimmunotherapy (RAIT). Long-term leukopenia (4-8 weeks) has been documented from a single maximum tolerated dose (MTD) in experimental mice. Several methods for alleviating RAIT-induced marrow toxicity have been evaluated preclinically, including cytokine intervention, bone marrow transplantation (BMT), and hemoregulatory peptide administration. To improve the therapeutic potential of RAIT, multiple doses of radioantibody must be delivered. Maximizing the frequency of radioantibody administration is desirable. However, little is known about the myelotoxic effects of multiple cycles of RAIT. In the studies presented here we compared the magnitude of myelosuppression, the time of nadir, and the duration of toxicity associated with one or two MTDs of 1-131-MN-14 anti-carcinoembryonic antigen immunoglobulin G (250 microCi) administered to BALB/c mice 49 days apart, the shortest interval possible without producing lethality. Studies were conducted with radioantibody alone or with cytokines (interleukin-1/granulocyte-macrophage colony-stimulating factor), BMT, or Hp5b to determine whether bone marrow became more "brittle" after the first dose. Profiles of myelosuppression and recovery were monitored weekly for 7 weeks after each dose in both granulocyte and lymphocyte populations. The results demonstrated that granulocyte suppression was greater and of longer duration after the second dose of RAIT administered alone, with cytokines, or with BMT, but less severe with Hp5b. For example, in the RAIT + BMT treatment, the second dose resulted in an 87% loss of granulocytes, whereas a 30% loss occurred after the first dose. The nadir of toxicity lasted until days 21 to 28 after the second dose and until day 14 after the first dose. Lymphocyte suppression was of greater duration after the second cycle of RAIT alone or RAIT with BMT, plateauing at <50% of untreated levels between days 28 and 49, but was of shorter duration when RAIT was given with cytokines or Hp5b. The results are discussed in terms of 1) the radiosensitivity of each subpopulation, 2) the effects on progenitors and on stromal cells, 3) the benefits of increasing dose frequency vs. increasing dose intensity, and 4) the possibility of using preclinical data to optimize the frequency of dosing in patient trials.