Precocious male puberty is a significant problem in Atlantic cod aquaculture. While photoperiod manipulation can inhibit testis growth, a detailed analysis of effects on spermatogenesis is missing. Starting July 1, 2004, prepubertal fish were exposed to different photoperiod regimens in indoor tanks for 17 mo. Testis histology, germ cell dynamics (proliferation and apoptosis), and plasma androgen levels were analyzed. In the natural light (NL) group, testis growth started in September 2004 and was completed in February 2005, when a 2-mo spawning period started. In the constant light (LL) group, none or very few spermatogenic cysts were recruited into spermatogenesis, and apoptotic germ cell loss was high. A change of photoperiod from NL to LL at winter solstice (December 21, 2004) resulted in premature (2 mo) completion of the reproductive cycle, while changing from LL to NL at winter solstice triggered faster than normal testis development. Plasma testosterone levels increased in the NL group from spermatogonial proliferation toward meiosis, while those of 11-ketotestosterone increased toward spermiogenesis and spermiation. Plasma androgen levels did not rise under LL conditions. Comparing fish with developing testes from all groups indicated that low androgen levels were associated with a high incidence of spermatogonial apoptosis; we also found that androgen receptor mRNA expression was most prominent in Sertoli cells in contact with growing spermatogonial clones. Our data show that an inhibitory photoperiod (LL) reduced or blocked differentiation of spermatogonia, increased apoptosis (particularly among proliferating spermatogonia), and was associated with reduced androgen levels, a situation possibly reflecting insufficient gonadotropic stimulation.