Our objective was to determine if rebreathing would reduce the gradient between arterial and end-tidal CO2 tension during positive-pressure ventilation.Experimental investigation.Anesthesiology laboratory.A total of 10 dogs of either sex.Anesthesia (sodium pentobarbital) and muscle relaxation (pancuronium) were induced and animals were tracheally intubated and ventilated with a standard anesthesia ventilator and breathing circuit with CO2 absorber and then with a Mapleson D circuit with a fresh gas flow rate (VF) equal to alveolar ventilation plus the sampling flow rate of two capnometers. Rebreathing was varied by adjusting the respiratory rate (RR) so that minute ventilation (VE) to VF ratio was 1:1, 2:1, 3:1, and 4:1.CO2 production (ATPD) was determined as the product of expired concentration of CO2 and VE (BTPS). Alveolar ventilation (VA) was calculated by dividing the product of CO2 production and barometric pressure corrected for ambient temperature and water vapor pressure at body temperature by PaCO2. Tidal volume, RR, airway gas temperature, concentration of CO2 in gas at the tracheal tube and inlet/outlet of the mechanical ventilator, body temperature, arterial gas tensions and pH, heart rate, arterial blood pressure, and cardiac output were measured. Minute ventilation, mean arterial blood pressure and end-expiratory CO2 tension (PECO2) (BTPS) were calculated. During positive-pressure ventilation, concentration of inspired CO2 was zero with standard circuitry, and significantly increased with Mapleson D when VE:VF ratio was 1:1 (0.56 +/- 0.19%), 2:1 (1.97 +/- 1.30%), 3:1 (2.56 +/- 1.05%), and 4:1 (3.01 +/- 1.45%) (p < 0.05). PECO2 was 34.8 +/- 3.2 mm Hg during ventilation with the standard circuit, and significantly increased during ventilation with Mapleson D when VE:VF ratio was increased from 1:1 (35.4 +/- 2.5 mm Hg) to 2:1 (40.2 +/- 3.6 mm Hg) and was not further increased at a VE:VF ratio of 3:1 (41.8 +/- 2.7 mm Hg) or 4:1 (41.3 +/- 2.4 mm Hg). The selected fresh gas flow rate was appropriate, because PaCO2 remained unchanged regardless of VE:VF ratio, indicating PaCO2 was dependent on VF, not on VE. The gradient between PaCO2 and PECO2 during ventilation with the standard circuit was 6.6 +/- 3.0 mm Hg; during ventilation with Mapleson D, it decreased significantly when VE:VF ratio was increased from 1:1 (6.5 +/- 3.6 mm Hg) to 2:1 (2.9 +/- 1.5 mm Hg), but was not significantly reduced further at 3:1 (1.7 +/- 1.1 mm Hg) or 4:1 (1.8 +/- 0.5 mm Hg) (p < 0.05).Rebreathing with a Mapleson D circuit and a VF equal to VA permitted normal CO2 elimination. Arterial PCO2 to PECO2 gradient decreased significantly during rebreathing, thus improving the reliability of capnography for estimating arterial PCO2. Consideration should be given to using the Mapleson D as a rebreathing circuit.