We tested in Siberian hamsters the nature of the maternal signal that relays photoperiodic information to the developing fetuses. As previous investigations have identified maternal hormonal and circadian components in this process, the specific goal of this presentation is to determine quality of the signal that connotes daylength when it is imparted to the fetus. Does the function of the signal received by the fetus best support the coincidence or duration hypotheses of photoperiodic induction? Pregnant hamsters received 1 or 8 hr melatonin or vehicle infusions everyday. Juveniles of intact mothers gestated on 16 hr of light per day (16L) experienced maximal suppression of testicular development when reared on 14L. However, when intact mothers gestated on 10L received a 1-hr melatonin infusion daily at 20:00-21:00 hr, their young responded to 14L with greatly accelerated testicular development. In the absence of the maternal pineal gland (and, therefore, the maternal melatonin signal), the effects of maternal melatonin infusions were reversed. Here, only the juveniles of 16L-gestated females infused at 20:00-21:00 hr daily responded to 14L with enhanced testicular development. All other groups showed the same extent of gonadal development, independent of the time or type of infusion their mothers received. Testicular development on 14L of all juveniles from pinealectomized mothers gestated on 10L was of the same magnitude, regardless of the type and time of infusion their mothers received during pregnancy. The results suggest that the maternal signal transferred to the fetuses during gestation consists not only of the daily melatonin signal, but also some circadian-based component that greatly affects the effect of the former. The timing, and not the duration, of the maternal melatonin signal with respect to the animals' (mother and fetus) circadian day is of crucial importance in the transfer of photoperiodic information from mother to fetus.