The authors compared PaCO2 with end-tidal CO2 (ETCO2) sampled at multiple sites along the endotracheal tube (ETT) in seven anesthetized rabbits (weight, 2.7-3.6 kg) to determine the most convenient, yet accurate, sampling location. Comparisons were made during spontaneous and controlled ventilation with fresh gas flows (FGF) of two and ten times the minute ventilation using a Mapleson D circuit. An Engstrom Eliza analyzer with a continuous sampling rate of 100 ml/min was used to measure ETCO2. A 0.75-mm ID polyethylene tube inserted in the side of the ETT sampled ETCO2 at the distal tip and at the 6-, 12-, and 15-cm marks on the ETT. ETCO2 was also measured at the standard proximal connector. The differences (P less than 0.05) between PaCO2 and ETCO2 at the distal, 6-, 12-, and 15-cm marks were 2.9 +/- 0.4, 3.1 +/- 0.4, 3.6 +/- 0.4, and 4.6 +/- 0.5 mmHg (mean +/- SEM), respectively, and did not change with FGF or mode of ventilation. The difference between PaCO2 and ETCO2 measured at the proximal connector was always large but significantly (P less than 0.05) greater during spontaneous than controlled ventilation (24.2 +/- 1.5 versus 15.0 +/- 1.4 mmHg) and at higher FGF (19.4 +/- 1.3 versus 16.8 +/- 1.6 mmHg). The differences (P less than 0.05) between ETCO2 at the distal tip and ETCO2 at the 6-, 12-, and 15-cm marks were 0.24 +/- 0.07, 0.73 +/- 0.11, and, 1.77 +/- 0.20 mmHg, respectively. This demonstrates that the change in ETCO2 between the distal tip and the 12-cm mark on the ETT is less than 1 mmHg, and that this clinically insignificant difference is independent of FGF and mode of ventilation. The 12 cm-mark is outside of the mouth on a newborn, and sampling ETCO2 at that point, which may be accomplished simply by inserting a small needle in the side of the ETT, may be the most appropriate sampling location.