BACKGROUND: Studies comparing intraoperative warming devices report discordant or out-of-date results. This trial compared two underbody warming devices. METHODS: Thirty patients undergoing elective prolonged laparoscopic gynecologic surgery were randomized to underbody warming by forced air (n = 15) or contact with a carbon fiber blanket (n = 15). The main outcome was esophageal temperature at the end of surgery. We also compared temperature throughout surgery and need for rescue warming, blood loss, fluids infused, urine output, and adverse events. Outcomes were compared with χ² or Fisher exact tests, t-tests, and mixed effects models as appropriate. RESULTS: No median (interquartile range) differences between forced-air and contact warming were found in initial or final temperatures (36.2 °C [36.0, 36.2] vs 36.3 °C [35.9, 36.6] and 36.6 °C [36.2, 36.8] vs 36.3 °C [35.6, 36.5]). Temperature slightly increased over time in the forced-air group and slightly decreased in contact group (0.11 °C/h (0.02, 0.19) vs -0.05 °C/h (-0.13, 0.03), p = 0.008). A single patient required rescue warming (in contact group after 4.75 h). Surgery took longer in the contact group (3.2 h [2.5, 3.8] vs 4.0 h [2.9, 5.6] h, p = 0.042). Two surgeons complained of dizziness related to ambient heat in the forced-air group. No differences were found in the remaining variables. CONCLUSIONS: During use of the underbody forced-air and carbon-fiber warming devices tested, we recorded only very slight differences in temperature changes over the course of surgery. The variations can be considered clinically unimportant as no significant difference was evident at the end of surgery.