To evaluate the measurement of cardiac output (CO) using continuous electrical bioimpedance cardiography (Physioflow; Neumedx, Philadelphia, PA) (CO(PF)) with a simultaneous direct Fick measurement (CO(FICK)) in children with congenital heart disease. A prospective cohort study comparing 2 methods of measurement of CO. A quaternary university-affiliated pediatric hospital. Children undergoing cardiac catheterization for clinical care. The Physioflow measured continuous real time CO in 15-second epochs and simultaneous measurement of cardiac output by direct Fick (with mass spectrometry to assess VO(2)) were acquired. Sixty-five patients were recruited, and data from 56 (25 males) were adequate for analysis. The median age at study was 3.5 years (range, 0.4-16.6 years), and the median body surface area was 0.62 m(2) (range, 0.31-1.71). There were 25 of 56 (45%) with univentricular physiology. A total of 19,228 Physioflow data points were available for the analysis of which 14,569 (76%) were valid; 96% of the invalid measurements were identified as artifacts by the device. The average cardiac index of valid measurements was 3.09 ± 0.72 L/min/m(2). Compared with the Fick CO, the mean bias was -0.09 L/min, but the 95% limits of agreement were -3.20 to +3.01 L/min/m(2). Consequently, only 20 of 56 (36%) of measurements were within 20%, and 31 of 56 (55%) of measurements were within 30% of each other. Compared with measurements made by direct Fick, CO measured using the Physioflow device was unreliable in anesthetized children with congenital heart disease.