OBJECTIVE: To determine whether end-tidal CO(2) (Petco(2)) measurement provides a reliable estimate of ventilation in critically ill children who are mechanically ventilated. DESIGN: Prospective, nonrandomized, consecutive enrollment study. SETTING: A university-affiliated children's hospital pediatric intensive care unit. PATIENTS: All intubated, mechanically ventilated pediatric patients. INTERVENTIONS: All Petco(2)-Paco(2) pairs were from patients ventilated with a Servo 300 Ventilator (Siemens-Elema AB, Stockholm, Sweden). When a blood gas sample was obtained, Petco(2) as measured by a continuous mainstream Petco(2) capnograph was recorded. Measurements: The results of blood gas measurements and corresponding Petco(2) measurements were recorded. Demographic data and primary diagnosis were noted. Petco(2)-Paco(2) pairs obtained from patients with intracardiac shunts or obtained during high-frequency oscillation or extracorporeal membrane oxygenation at the time of measurement were excluded from analysis. Linear regression was used to analyze Petco(2)-Paco(2) pairs. Repeated measure analysis of variance with the mixed-model algorithm in SAS software (SAS Institute, Carey, NC) was used to analyze the trend in the Petco(2) and Paco(2) relationship. Chi-square was used to analyze categorical data. Statistical significance was considered p <.05. RESULTS: A total of 129 children were enrolled, and 1708 paired Paco(2) and Petco(2) measurements were recorded. The mean age +/- sd was 4.1 +/- 5.6 yrs. Paco(2) positively correlated with Petco(2). The linear equation for the regression analysis was y = 0.71x (95% confidence interval, 0.69-0.73) + 8.93 (95% confidence interval, 7.89-9.97), with r (2) =.716 and p <.001. The Petco(2)-Paco(2) difference was </=5 mm Hg (0.67 kPa) in 54% and </=10 mm Hg (1.33 kPa) in 80% of paired data. Increased lung disease had a negative effect on Petco(2) correlation with Paco(2). A total of 223 of 640 (35%) blood gases (defined by Pao(2)/Fio(2) ratio of <200) had >10 mm Hg (1.33 kPa) difference between the Petco(2) and Paco(2). However, only 111 of 1068 (10%) Petco(2)-Paco(2) pairs had a difference of >10 mm Hg (1.33 kPa) in patients with a Pao(2)/Fio(2) ratio >200. Trend analysis showed the Petco(2)-Paco(2) difference increased with increasing duration of mechanical ventilation. CONCLUSION: In most intubated, mechanically ventilated infants and children, Petco(2) reliably estimates ventilation.