Sublethal cyclophosphamide treatment induces unique regeneration patterns in bone marrow and the spleen of a mouse. Colony-forming units spleen (CFU-S)(day 8), CFU-granulocyte-macrophage (GM), nucleated cell counts, and their differentials in bone marrow, spleen, and peripheral blood were determined in mice treated with a single dose of cyclophosphamide. To study further the mechanisms underlying the unique patterns of hematopoietic regeneration after cyclophosphamide, mRNA levels for stem cell factor (SCF), Flt-3 ligand, and macrophage inflammatory factor (MIP)-1 alpha cytokines were determined in bone marrow and spleen. Granulocyte precursor cells were less depleted by cyclophosphamide compared to erythroid nucleated cells and lymphocytes both in bone marrow and spleen. Rapid expansion of granulopoietic cells increased the granulocytic/erythroid ratio significantly during regeneration. CFU-S in the bone marrow and the spleen showed different sensitivity in vivo but not in vitro to cyclophosphamide; CFU-GM were equisensitive in both sites. In bone marrow, an initial fast recovery of CFU-S and CFU-GM on days 2 to 3 was followed by a secondary deep decline in their numbers occurring between days 5 and 7. This decline was accompanied with a depression of CFU-S proliferation and with significantly increased CFU-S numbers in the peripheral blood. In the spleen, absolute CFU-S and CFU-GM numbers were increased several-fold at this time. Seven days after cyclophosphamide, the spleen contained 69% of the total body CFU-S compared to 4% in controls. Splenectomy did not abolish the secondary disease of CFU-S in the bone marrow, but it led to a marked elevation of circulating leukocytes and CFU-S. There was an eight-fold increase in the SCF mRNA level in the bone marrow 2 days after cyclophosphamide, corresponding with a high proliferation rate of CFU-S. No significant changes in mRNAs for Flt-3 ligand and MIP-1 alpha have been found. This in-depth analysis of murine hematopoietic responses to cyclophosphamide provides evidence for the complexity of the involved local and systemic regulations. This represents a significant challenge to experimental hematology, which could now be tackled with methods allowing the study of changes in the gene expression during cyclophosphamide-induced hematopoietic damage.