Spermatogenesis is regulated by hormones, local regulatory factors in the testes and specific gene expression of spermatogenic cells in humans. In this study, we have detected the expression of the deleted in azoospermia (DAZ), the DAZ-like autosome (DAZL1), and the protamine-2 genes in spermatogenic cells. Spermatogenesis in 38 male infertility patients was evaluated by the semen analysis and histological examination. Patients were diagnosed as Sertoli cell-only syndrome (n = 20), maturation arrest (n = 6), hypospermatogenesis (n = 6), and obstructive azoospermic patients with normal spermatogenesis (n = 6). After microscopic observation of the wet preparation of the testis tissues, seminiferous tubule contents were used for reverse transcription-polymerase chain reaction (RT-PCR) analysis of DAZ, DAZL1 and protamine-2. In cases with Sertoli-cell only syndrome, we found spermatogenic cells in 30% of patients (6/20) by the wet preparation method. There was no difference between the histology and the wet preparation results in maturation arrest and obstructive azoospermia; however, in one case of hypospermatogenesis, spermatozoa were not detectable by the wet preparation method. Using in-situ hybridization with DAZ and protamine-2 ribonuclear probes, we confirmed spermatogenic cell-specific expression of DAZ (spermatogonia/early spermatocyte) and protamine-2 (spermatid/spermatozoon). DAZ and protamine-2 expression can therefore be considered spermatogenic cell markers and could be useful in molecular diagnosis of spermatogenesis. In 13 patients with spermatozoa under the wet preparation, the expression of DAZ, DAZL1 and protamine-2 was detected in all the preparations. In one wet preparation showing only spermatogonia/spermatocyte, only DAZ and DAZL1 RNA were detected. In 14 wet preparations showing no spermatogenic cells, DAZ, DAZL1 and protamine-2 were not detected except in one preparation where DAZL1 expression was detected. In 10 wet preparations representing spermatogonia/spermatocyte to spermatids, but showing no spermaozoa, DAZ and DAZL1 were detected in eight and nine preparations respectively, and protamine-2 was detected in six preparations. These results of gene expression were similar to the wet preparation results. RT-PCR for DAZ, DAZL1 and protamine-2 was informative for the existence of germ cells, germ cell physiology and differentiation. From these results, we suggest that the analysis of DAZ, DAZL1 and protamine-2 expression by RT-PCR and wet preparation might offer a better method for finding the spermatogenic cells compared to the histological method.