BACKGROUND: Passive humidifiers have gained acceptance in the intensive care unit because of their low cost, simple operation, and elimination of condensate from the breathing circuit. However, the additional dead space of these devices may adversely affect respiratory function in certain patients. This study evaluates the effects of passive humidifier dead space on respiratory function. METHODS: Two groups of patients were studied. The first group consisted of patients recovering from acute lung injury and breathing spontaneously on pressure support ventilation. The second group consisted of patients who were receiving controlled mechanical ventilation and were chemically paralyzed following operative procedures. All patients used 3 humidification devices in random order for one hour each. The devices were a heated humidifier (HH), a hygroscopic heat and moisture exchanger (HHME) with a dead space of 28 mL, and a heat and moisture exchanger (HME) with a dead space of 90 mL. During each measurement period the following were recorded: tidal volume, minute volume, respiratory frequency, oxygen consumption, carbon dioxide production, ratio of dead space volume to tidal volume (VD/VT), and blood gases. In the second group, intrinsic positive end-expiratory pressure was also measured. RESULTS: Addition of either of the passive humidifiers was associated with increased VD/VT. In spontaneously breathing patients, VD/VT increased from 59 +/- 13 (HH) to 62 +/- 13 (HHME) to 68 +/- 11% (HME) (p < 0.05). In these patients, constant alveolar ventilation was maintained as a result of increased respiratory frequency, from 22.1 +/- 6.6 breaths/min (HH) to 24.5 +/- 6.9 breaths/min (HHME) to 27.7 +/- 7.4 breaths/min (HME) (p < 0.05), and increased minute volume, from 9.1 +/- 3.5 L/min (HH) to 9.9 +/- 3.6 L/min (HHME) to 11.7 +/- 4.2 L/min (HME) (p < 0.05). There were no changes in blood gases or carbon dioxide production. In the paralyzed patient group, VD/VT increased from 54 +/- 12% (HH) to 56 +/- 10% (HHME) to 59 +/- 11% (HME) (p < 0.05) and arterial partial pressure of carbon dioxide (PaCO2) increased from 43.2 +/- 8.5 mm Hg (HH) to 43.9 +/- 8.7 mm Hg (HHME) to 46.8 +/- 11 mm Hg (HME) (p < 0.05). There were no changes in respiratory frequency, tidal volume, minute volume, carbon dioxide production, or intrinsic positive end-expiratory pressure. DISCUSSION: These findings suggest that use of passive humidifiers with increased dead space is associated with increased VD/VT. In spontaneously breathing patients this is associated with an increase in respiratory rate and minute volume to maintain constant alveolar ventilation. In paralyzed patients this is associated with a small but statistically significant increase in PaCO2. CONCLUSION: Clinicians should be aware that each type of passive humidifier has inherent dead space characteristics. Passive humidifiers with high dead space may negatively impact the respiratory function of spontaneously breathing patients or carbon dioxide retention in paralyzed patients. When choosing a passive humidifier, the device with the smallest dead space, but which meets the desired moisture output requirements, should be selected.