1. Rabbits were clipped and exposed in turn to three environmental conditions: control (C), cold exposure (CE) and water deprivation (WD). Following each type of treatment, the rabbits were exposed to an ambient temperature (Ta) of 35 degrees C for 1 hr. Throughout this period they breathed either normal atmospheric air or 6% CO2 in air. 2. During heat exposure, measurements were made of the respiratory responses and of the O2 consumption (Vo2) of the rabbits. Rectal temperature (Tre) was measured immediately before and again immediately after heat exposure. 3. When subjected to cold exposure or water deprivation the rabbits showed an initial decrease in respiratory frequency (RF) and an initial increase in VT when compared with controls. There was no difference in VE. Rabbits breathing 6% CO2 showed an increase in VT and VE and a decrease in RF when compared with rabbits breathing atmospheric air. In all cases a change in VT or RF was associated with a reciprocal change in the other parameter. 4. The respiratory responses to breathing 6% CO2 were essentially similar in treated and control rabbits, from which it is concluded that neither cold exposure nor water deprivation alter the sensitivity of the medullary respiratory centre to the respiratory drive from the central chemosensors. 5. The increase in Tre during heat exposure was significantly less in rabbits breathing 6% CO2 than in rabbits breathing atmospheric air. However, there was no significant over-all difference in VO2 between rabbits breathing CO2 and those breathing air. From this it is concluded that increased ventilation induced by CO2 causes a greater dissipation of heat than does thermally-induced panting. 6. It is concluded that VT is controlled by the level of blood PCO2 whereas RF is controlled by thermoregulatory requirements. It is further concluded that the reciprocal relationship between VT and RF is regulated in such a way as to maintain VE at the appropriate level for effecting gaseous exchange and evaporative heat loss.