The vertebrate gonad consists of two cell populations: germ cells (GCs), which produce the gametes, and somatic gonad cells (SGCs), which serve two important functions. First, SGCs create an environment that protects GCs and nurtures their development through cell signaling. Second, a subset of SGCs secrete hormones required for secondary sexual development, and therefore defects in SGC development result in disorders of sexual development. Despite the known importance of SGCs for fertility, little is known how the early somatic gonad forms or which genes are required for its formation. We have discovered that a mutation in the zebrafish Fibroblast growth factor (Fgf) ligand fgf24 results in a reduction of GCs during larval development, at a time that precedes sexual differentiation. The phenotype persists into adulthood, and most mutant adults are sterile. Upon examination of the 10 days post fertilization (dpf) gonad by transmission electron microscopy (TEM), we found that wild-type fish have multilayered gonads, where the SGCs have already begun to separate into two distinct layers and the GCs are arranged in the center, protected by the SGCs. We have determined that fgf24 is expressed in the outer layer of epithelial-like SGCs. In contrast, a known Fgf-responsive gene, ets variant 4 (etv4/pea3), is expressed in the population of mesenchymal-like SGCs more internally located. Based on the TEM data and Laminin immunostaining, we have found that these layers are separated by a basement membrane. In contrast, the gonads of fgf24 mutants have only one layer of SGCs enveloping the GCs, lack a basement membrane, and fail to express etv4. Furthermore, we asked whether genes known to be important for SGC function were expressed in fgf24 mutants: gata4, a transcription factor required for mouse gonadogenesis; cyp19a1a, an aromatase that converts androgens to estrogens; and amh, a TGF-β ligand required for male sexual development in many vertebrates. In accordance with our hypothesis, in situ hybridization showed that these genes were expressed in the mesenchymal SGC population of wild-type larval fish, but their expression was greatly reduced or completely absent in fgf24 mutant gonads. These results support our overall hypothesis that Fgf24 functions primarily to promote development of the mesenchymal SGCs and overall morphogenesis of the somatic gonad, and that the loss of GCs in fgf24 mutants is a secondary consequence of a somatic gonad defect.