To determine the accuracy of end-tidal PCO2 (PETCO2) measurements analyzed with a sidestream capnometer in infants and children whose lungs were ventilated with a Sechrist infant ventilator and an Ayre's t-piece, we compared PETCO2 measurements obtained from the proximal (PETCO2-p) and distal (PETCO2-d) ends of the tracheal tube to arterial PCO2 (PaCO2) in 37 healthy infants and children between 1.3 and 24.5 kg. Both PETCO2-p and PETCO2-d accurately approximated PaCO2, however, the mean (+/- SD) arterial to end-tidal PCO2 difference (delta(a-ET)PCO2) was significantly greater with proximal (1.27 +/- 1.54 mmHg) than with distal sampling (0.64 +/- 1.64 mmHg) (P less than 0.01). In the subgroup of patients who weighted less than 12 kg, the delta(a-ET)PCO2 using proximal gas sampling (1.94 +/- 1.29 mmHg) was also significantly greater than it was using distal sampling (0.74 +/- 1.31 mmHg) (P less than 0.001). We conclude that although statistically different, both proximal and distal estimates of PETCO2 provide acceptable estimates of PaCO2 in healthy infants and children who are ventilated with a Sechrist infant ventilator and an Ayre's t-piece system.