The use of a transthoracic bioimpedance monitor to determine cardiac output was evaluated in critically ill children. The children ranged in age from 10 months to 8 yr and their height and weight ranged from the third to the 97th percentile. Each child had a thermodilution catheter in place to monitor cardiac output. The bioimpedance monitor used in this study, the NCCOM-3, required the input of a constant (L), which was obtained for each individual patient by adjusting the L setting until cardiac output measured by bioimpedance (COBI) was within 10% of cardiac output measured by thermodilution (COTD). This method of determining L was superior to using either measured thoracic length or the manufacturer's guidelines to obtain L and resulted in an excellent correlation between COTD and COBI (r = .94; p less than .05; n = 59). In children less than 125 cm in height, measured thoracic length alone was inadequate to use for L but provided a good approximation of L when multiplied by 1.25. This study suggests that the use of transthoracic bioimpedance to determine cardiac output compares favorably with thermodilution techniques and it is noninvasive.