Processing of pre-mRNA into mRNA is an important regulatory mechanism in eukaryotes that is mediated by the spliceosome, a huge and dynamic ribonucleoprotein complex. Splicing defects are implicated in a spectrum of human disease, but the underlying mechanistic links remain largely unresolved. Using a genome-wide association approach, we have recently identified single nucleotide polymorphisms in humans that associate with nonobstructive azoospermia (NOA), a common cause of male infertility. Here, using genetic manipulation of corresponding candidate loci in Drosophila, we show that the spliceosome component SNRPA1/U2A is essential for male fertility. Loss of U2A in germ cells of the Drosophila testis does not affect germline stem cells, but does result in the accumulation of mitotic spermatogonia that fail to differentiate into spermatocytes and mature sperm. Lack of U2A causes insufficient splicing of mRNAs required for the transition of germ cells from proliferation to differentiation. We show that germ cell-specific disruption of other components of the major spliceosome manifests with the same phenotype, demonstrating that mRNA processing is required for the differentiation of spermatogonia. This requirement is conserved, and expression of human SNRPA1 fully restores spermatogenesis in U2A mutant flies. We further report that several missense mutations in human SNRPA1 that inhibit the assembly of the major spliceosome dominantly disrupt spermatogonial differentiation in Drosophila. Collectively, our findings uncover a conserved and specific requirement for the major spliceosome during the transition from spermatogonial proliferation to differentiation in the male testis, suggesting that spliceosome defects affecting the differentiation of human spermatogonia contribute to NOA.