The duration of nocturnal pineal melatonin secretion transduces effects of day length (DL) on the neuroendocrine axis of photoperiodic rodents. Long DLs support reproduction, and short DLs induce testicular regression, followed several months later by spontaneous recrudescence; gonadal regrowth is thought to reflect development of tissue refractoriness to melatonin. In most photoperiodic species, pinealectomy does not diminish reproductive competence in long DLs. Turkish hamsters (Mesocricetus brandti) deviate from this norm: elimination of melatonin secretion in long-day males by pinealectomy or constant light treatment induces testicular regression and subsequently recrudescence; the time course of these gonadal transitions is similar to that observed in males transferred from long to short DLs. In the present study, long-day Turkish hamsters that underwent testicular regression and recrudescence in constant light subsequently were completely unresponsive to the antigonadal effects of short DLs. Other hamsters that manifested testicular regression and recrudescence in short DLs were unresponsive to the antigonadal effects of pinealectomy or constant light. Long-term suppression of melatonin secretion induces a physiological state in Turkish hamsters similar or identical to the neuroendocrine refractoriness produced by short-day melatonin signals (i.e., neural refractoriness to melatonin develops in the absence of circulating melatonin secretion). A melatonin-independent interval timer, which would remain operative in the absence of melatonin during hibernation, may determine the onset of testicular recrudescence in the spring. In this respect, Turkish hamsters differ from most other photoperiodic rodents.