STUDY OBJECTIVE: To compare the thermal and humidification capacity of three heated hot water systems (HHWSs) and two heat and moisture exchangers (HMEs) in ICU patients submitted to minute ventilation > 10 L/min. DESIGN: Prospective, controlled, randomized, not blinded study. SETTING: ICU of a university hospital. PATIENTS: ICU patients requiring controlled mechanical ventilation with minute ventilation > 10 L/min. Patients had to be sedated and paralyzed and had to require ventilation for more than four days. INTERVENTIONS: Following a randomized order, the patients were ventilated for 24-h periods with three HHWSs (Bennett Cascade 2 humidifier, Fisher-Paykel MR 460 and MR 600) and two HMEs (Pall Ultipor and Hygrobac filter). MEASUREMENTS AND RESULTS: In each patient and for each 24-h period, absolute humidity, (AH), relative humidity (RH) of inspired gases, and tracheal temperature were obtained. Two HHWSs (Bennett and Fisher-Paykel MR 460) had a better thermal and humidification capacity than any other systems (p < 0.001). The hydrophobic HME (Pall filter) had a poor thermal and humidification capacity (RH: 79 +/- 8.7 percent; AH: 20.6 +/- 2.3 mg H2O/L). The hygroscopic filter (Hygrobac filter) had better thermal and humidification capacity than the Pall filter (RH: 92.5 +/- 3.6 percent; AH: 29.1 +/- 1.8 mg H2O/L; p < 0.001). Tracheal temperature was well preserved by all systems. The thermic and humidification capability of the Hygrobac filter declined over 24 h. Since the Pall filter could not achieve an AH > 25 mg H2O/L in any patient, it was not studied beyond the first measurement. CONCLUSIONS: The Hygrobac filter had a thermal and humidification capability closed to the two HHWSs (81 to 97 percent) but the capability declines over 24 h. The Pall filter had a poor capability (54 to 74 percent of that of HHWSs).