Experiments in vitro suggest that although interleukin 5 (IL-5) stimulates the late stages of eosinophil differentiation, other cytokines are required for the generation of eosinophil progenitor cells. In this study transgenic mice constitutively expressing the IL-5 gene were established using a genomic fragment of the IL-5 gene coupled to the dominant control region from the gene encoding human CD2. Four independent eosinophilic transgenic lines have thus far been established, two of which with 8 and 49 transgene copies, are described in detail. These mice appeared macroscopically normal apart from splenomegaly. Eosinophils were at least 65- and 265-fold higher in blood from transgenics, relative to normal littermates, and approximately two- or sevenfold more numerous relative to blood from mice infected with the helminth Mesocestoides corti. Much more modest increases in blood neutrophil, lymphocyte, and monocyte numbers were noted in transgenics, relative to normal littermates (less than threefold). Thus IL-5 in vivo is relatively specific for the eosinophil lineage. Large numbers of eosinophils were present in spleen, bone marrow, and peritoneal exudate, and were highest in the line with the greatest transgene copy number. Eosinophilia was also noted in histological sections of transgenic lungs, Peyer's patches, mesenteric lymph nodes, and gut lamina propria but not in other tissues examined. IL-5 was detected in the sera of transgenics at levels comparable to those seen in sera from parasite-infected animals. IL-3 and granulocyte/macrophage colony-stimulating factor (GM-CSF) were not found. IL-5 mRNA was detected in transgenic thymus, Peyer's patches, and superficial lymph nodes, but not in heart, liver, brain, or skeletal muscle or in any tissues from nontransgenics. Bone marrow from transgenic mice was rich in IL-5-dependent eosinophil precursors. These data indicate that induction of the IL-5 gene is sufficient for production of eosinophilia, and that IL-5 can induce the full pathway of eosinophil differentiation. IL-5 may therefore not be restricted in action to the later stages of eosinophil differentiation, as suggested by earlier in vitro studies.