OBJECTIVE To compare the effects of two external rewarming methods on body core temperature and the rate of rewarming between two age groups (less than 65 years, 65 years or more) of adult, mildly hypothermic patients who have undergone cardiac surgery, during the immediate postoperative period. METHODS Stratified, randomized clinical trial. METHODS Five-bed cardiac surgical intensive care unit in a large teaching-research institution. METHODS Thirty-two white patients who had undergone cardiac surgery and who had mildly hypothermic body core temperatures (33 degrees to 35 degrees C) immediately after the surgery. METHODS Body core temperature was measured with a pulmonary artery catheter thermistor at the time of external rewarming method application and at 60, 90, and 150 minutes afterward. Rate of rewarming was measured as body core temperature change in degrees Celsius per hour (at 36.6 degrees C, minus body core temperature when external rewarming method was applied, divided by total rewarming time). Temperatures were recorded six times at intervals of 15 minutes; then every 30 minutes until a value of 36.6 degrees C was obtained, at which time the blanket was removed; then hourly for 8 hours. METHODS Either a fluid-filled circulating blanket (active-conductive external rewarming) or a reflective blanket (passive-reflective external rewarming) was applied immediately after core temperature was measured on admission to the cardiac surgical intensive care unit after surgery. RESULTS External rewarming methods affected body core temperature differently at different times, and there were significant differences in body core temperature across the time periods (p < 0.05). Both active and passive external rewarming methods showed a sigmoidal rewarming pattern without a downward temperature drift. The fluid-filled circulating blanket produced a quicker and steeper body core temperature change in the early rewarming phase; the reflective blanket resulted in a more gradual temperature rise. Age did not significantly affect body core temperature, nor did age or external rewarming method significantly influence the rate of rewarming, although total rewarming time was longer for those of more advanced age. Seven subjects with passive rewarming method experienced body core temperature overshoot during the 8-hour period after blanket removal. CONCLUSIONS In this study, conduction and reflection of radiant heat were equally effective in producing an acceptable rate of rewarming but contributed to different internal patterns in core rewarming. The average total rewarming time with the active external rewarming method was 1 hour shorter than with the passive external rewarming method.