Spermatogenesis is maintained by a pool of spermatogonial stem cells (SSCs). Analyses of the molecular profile of SSCs have revealed the existence of subsets, indicating that the stem cell population is more heterogeneous than previously believed. However, SSC subsets are poorly characterized. In rodents, the first steps in spermatogenesis have been extensively investigated, both under physiological conditions and during the regenerative phase that follows germ cell damage. In the widely accepted model, the SSCs are type Asingle (As) spermatogonia. Here, we tested the hypothesis that As spermatogonia are phenotypically heterogeneous by analyzing glial cell line-derived neurotrophic factor (GDNF) family receptor alpha1 (GFRA1) expression in whole-mounted seminiferous tubules, via cytofluorimetric analysis and in vivo colonogenic assays. GFRA1 is a coreceptor for GDNF, a Sertoli cell-derived factor essential for SSC self-renewal and proliferation. Morphometric analysis demonstrated that 10% of As spermatogonia did not express GFRA1 but were colonogenic, as shown by germ cell transplantation assay. In contrast, cells selected for GFRA1 expression were not colonogenic in vivo. In human testes, GFRA1 was also heterogeneously expressed in Adark and in Apale spermatogonia, the earliest spermatogonia. In vivo 5-bromo-2'-deoxyuridine administration showed that both GFRA1(+) and GFRA1(-) As spermatogonia were engaged in the cell cycle, a finding supported by the lack of long-term label-retaining As spermatogonia. GFRA1 expression was asymmetric in 5% of paired cells, suggesting that As subsets may be generated by asymmetric cell division. Our data support the hypothesis of the existence of SSC subsets and reveal a previously unrecognized heterogeneity in the expression profile of As spermatogonia in vivo.