Liquid ventilation with perfluorocarbon (PFCV) has advantages over conventional gas ventilation (GV) in premature and lung-injured newborn animals. Indirect calorimetric measurement of both oxygen consumption (VO2) and carbon dioxide production (VCO2) during PFCV has not been previously performed. In addition, comparison to indirect calorimetric measurement of VO2 and VCO2 during GV has not been evaluated. Ten fasted normal cats weighing 2.6 to 3.9 kg were anesthetized with pentobarbital and pancuronium. Tracheostomy was performed. Gas exchange was measured across the native lung during GV and across the membrane lung of the liquid ventilator during PFCV. VO2 was measured using a modification of a previously described, indirect, closed-circuit, volumetric technique. VCO2 was analyzed by capnographic assay of the mixed-expired closed-circuit air. The VCO2/VO2 ratio (RQ) was calculated. There was no change in VO2, VCO2, or RQ during PFCV when compared with GV (VO2: GV = 5.7 +/- 0.3 mL/kg/min, PFCV = 5.6 +/- 0.5 mL/kg/min [P = NS]; VCO2: GV = 4.9 +/- 1.1 mL/kg/min, PFCV = 4.8 +/- 0.9 mL/kg/min [P = NS]; RQ: GV = 0.85 +/- 0.21, PFCV = 0.86 +/- 0.21 [P = NS]). During GV the PaO2 was higher than during PFCV (PaO2: GV = 335 +/- 70 mm Hg, PFCV = 267 +/- 83 mm Hg [P = .04]), as is expected because of the relative reduction in the inspiratory PiO2 of the perfluorocarbon during liquid ventilation.(ABSTRACT TRUNCATED AT 250 WORDS)