In many mammals the ventilatory response to hypoxia depends on ambient temperature (Ta), largely because of the hypometabolic effects of hypoxia below thermoneutrality. We questioned whether the ventilatory response to asphyxia also depends upon Ta, and the role played by metabolism and body temperature (Tb). Oxygen consumption (VO2) and pulmonary ventilation (VE) were measured in conscious rats at Ta = 27 degrees C (warm) and 11 degrees C (cold), breathing air or two levels of asphyxic gases, moderate (10% O2-4% CO2), or severe (10% O2-8% CO2), for approximately 30 min each. In the cold, the pattern of the VE response to moderate asphyxia was qualitatively similar to that seen in hypoxia alone, i.e the attained VE/VO2 was similar in warm and cold conditions, with, in the latter, a major drop in VO2 and little or no hyperpnea. During severe asphyxia, however, the VE/VO2 attained in the cold was less than in the warm, and it was accompanied by a large drop in Tb (approximately 6 degrees C). Blood gases confirmed the lower asphyxic hyperventilation in the cold. By maintaining Tb at 38 degrees C with an implanted abdominal heat exchanger, the VE/VO2 levels attained during asphyxia were the same between cold and warm conditions. We conclude that (a) the VE response to asphyxia is Ta-dependent, largely because of the hypometabolic effect of the hypoxic component in the cold, (b) during moderate asphyxia the hypercapnic component is qualitatively unimportant, and (c) with severe asphyxia the hypercapnia becomes an important contributor to the Ta-sensitivity by aggravating the decrease in Tb in the cold and lowering VE sensitivity.