We have performed a comprehensive analysis of cell lines and tissues to compare and contrast the expression patterns of Flt3 ligand (FL), c-Kit ligand (KL), and macrophage colony-stimulating factor as well as their receptors, Flt3, c-Kit, and c-Fms. The message for FL is unusually ubiquitous, whereas that of its receptor is quite restricted, apparently limiting the function of the ligand to fetal development and early hematopoiesis. We have also sequenced a mouse FL genomic clone, revealing how the three splice variant FL mRNAs that we have isolated arise. The chromosomal location of the FL gene has been mapped, by in situ hybridization, to chromosome 7 in mouse and chromosome 19 in human. Natural FL protein has been purified from a stromal cell line and shown to be a 65 kD nondisulfide-linked homodimeric glycoprotein comprised of 30 kD subunits, each containing 12 kD of N- and O-linked sugars. Pulse-chase experiments show that one of the splice variants (T110) is responsible for producing the bulk of soluble FL, but only after it has first been expressed at the cell surface as a membrane-bound form. The other splice-variant forms produce molecules that are either obligatorily soluble (T169) or membrane-bound but released only very slowly (T118). Finally, even though most cell lines express some amount of FL mRNA, we found that very little FL protein is actually made, with T cells and stromal cells being the major producers. The data suggests that FL plays its roles over very short distances, perhaps requiring cell-cell contact.