Variations in end-tidal carbon dioxide partial pressure (PETCO2) and temperature were measured for six hours following coronary artery bypass surgery in twenty patients. In the recovery room, the patients were mechanically ventilated with a tidal volume of 12 ml X kg-1. Arterial blood gases were drawn every two hours, and the respiratory frequency was adjusted to maintain arterial carbon dioxide pressure (PaCO2) in the range of 30-45 mmHg. Naso-pharyngeal temperature was recorded every 30 minutes, and PETCO2 was measured continuously. The mean difference between temperature-corrected arterial and end-tidal CO2 pressure measurements was 3.2 mmHg (SD = 2.8; r = 0.963). This difference did not vary with time, temperature or PCO2. The largest temperature increases (mean 1.7 degree C/hour) occurred at a mean of 253 minutes after the end of surgery. End-tidal PCO2 increased markedly as temperature rose, in spite of a coincident increase in ventilation and then decreased as temperature stabilized. Large increases in CO2 production, caused by the metabolic demands during rewarming, most likely account for these changes. It is concluded that end-tidal CO2 recordings are reliable, and can help in maintaining normocarbia during the short but unstable period associated with rewarming following cardiac surgery.