Study Objective: To compare passive heat retention by low-flow anesthesia, alone and with additional thermal insulation by reflective blankets, with forced-air warming preventing intraoperative hypothermia during combined epidural-general anesthesia. Design: Randomized, controlled study. Setting: Inpatient anesthesia at a university department of orthopedic surgery. Patients: 30 ASA physical status I and II patients, who were scheduled for elective hip or knee arthroplasty and were free from systemic disease. Interventions: Patients received epidural block up to T 10 by alkalinized lidocaine 2 %, and then were administered standard general anesthesia by means of low-flow rebreathing system (fresh gas flow = 1 L/min). All procedures started between 8 and 10 am, and operating room (OR) temperature was maintained between 21 ° and 23 °C, with relative humidity ranging between 40% and 45%. For heat retention or warming therapy, patients received either low-flow anesthesia only (control, n = 10), low-flow anesthesia with additional reflective blankets (blanket, n = 10), or low-flow anesthesia with active forced-air warming (forced-air, n = 10). Tympanic temperature was measured at OR arrival (baseline); immediately following general anesthesia induction; 30, 60, 90, and 120 minutes from general anesthesia induction; and at the end of surgery. Measurements and Main Results: Duration of anesthesia, invasiveness of surgery, and baseline core temperature were similar in the three groups. Core temperature decreased in all the three groups 30 minutes after general anesthesia induction compared with baseline ( p < 0.01); afterwards, it progressively decreased in the control and blankets groups ( p = 0.004), with a reduction from baseline values measured at the end of surgery of 2.0 °C and 1.6 °C, respectively. In the forced-air group, after the initial significant decrease ( p = 0.01 vs. baseline), core temperature progressively increased to 35.8 ± 0.6 °C, which was similar to preoperative values and significantly higher than either the control or blankets groups ( p = 0.004). Conclusions: During combined epidural-general anesthesia for elective hip and knee arthroplasty, passive heat retention by means of low-flow anesthesia alone and in combination with reflective blankets is ineffective in maintaining intraoperative normothermia and definitely inferior to active forced-air warming.