Individuals of many nontropical rodent species restrict breeding to the spring and summer. Seasonal reproductive quiescence putatively reflects the energetic incompatibility of breeding and thermoregulatory activities. However, so-called "out-of-season" breeding occurs in virtually all rodent populations examined, suggesting that the incompatibility can be resolved. Both reproductive inhibition and development of energy-saving adaptations are mediated by environmental photoperiod, but some individuals do not inhibit reproduction in short days. In order to assess the costs and benefits of winter breeding, the present study examined the extent to which male prairie voles (Microtus ochrogaster) and deer mice (Peromyscus maniculatus) that maintained summer reproductive function in winter-simulated daylengths also maintained summer thermoregulatory adaptations. Circadian locomotor activity patterns, basal metabolic rate, capacity for nonshivering thermogenesis, nest building, body mass, and daily food consumption were compared among short-day (LD 8:16) regressed males, short-day (LD 8:16) nonregressed males, and long-day (LD 16:8) males. Short-day nonregressed deer mice resembled long-day conspecifics in terms of body mass and nest-building activities; however, the locomotor activity pattern of short-day nonregressed deer mice was similar to that of their short-day regressed conspecifics. Short-day nonregressed prairie voles had body masses similar to those of long-day conspecifics. Regardless of their reproductive response to photoperiod, short-day prairie voles reduced their daily food consumption and wheel-running activity, compared to long-day voles. These results suggest that winter breeding has energetic costs, most likely resulting from maintaining a "summer-like" body mass relative to that of reproductively regressed animals. These costs may be ameliorated to some extent by the reduction in locomotor activity and nest-building behavior emitted by short-day animals, regardless of reproductive response to short days. Thus, the occurrence of winter breeding may be the result of sufficient numbers of reproductively photoperiod-nonresponsive morphs in the population and sufficiently mild ambient conditions to permit survival of these larger animals.