Many studies point out that ambient temperatures outside the thermoneutral zone disturb sleep. Thermal responses depend on the phase of sleep and it appears that rapid eye movement (REM) sleep and thermoregulation are mutually exclusive in animals. In humans, there is evidence that this thermoregulatory disruption is less important since the thermal responses to heat or cold exposures are not abolished. Thermal responses persist during slow wave sleep (SWS). These differences suggest that there is a hierarchical control of body temperature regulation. The control includes different integrators at many levels of the nervous system with each level facilitated or inhibited by levels above and below. During SWS, body temperature is controlled by diencephalic structures whereas during REM sleep there is a suspension of hypothalamic thermoregulatory influences. The thermoregulatory differences may be accounted for by the neural organization and the autonomy of brain stem and spinal operative levels which differ between sleep phases and species. However, in the neonate, thermoregulation is not impaired during active sleep (AS) and Quiet Sleep (QS) considered, from behavioral and polygraphic measures, as immature forms of REM and SWS, respectively. From a thermoregulatory point of view, AS seems to be a well-protected sleep phase. As regards the thermal responses, these observations suggest that neonatal active sleep may be a precursor both of REM sleep and SWS in adults.