This study compares the correlation between two methods for the determination of cardiac output-the pulmonary arterial thermodilution technique using the Swan-Ganz catheter and the femoral arterial thermodilution technique using a pulse contour analysis computer (PiCCO) catheter. We performed a prospective animal study using 16 immature Maryland pigs weighing 9 to 16 kg. A 5.5- or 7.5-Fr Swan-Ganz catheter was introduced into the femoral or jugular vein, and a 4- or 5-Fr arterial PiCCO catheter was introduced into the femoral artery. In each animal, we made measurements of cardiac output at 30-minute intervals, simultaneously by pulmonary arterial thermodilution and femoral arterial thermodilution, before, during, and after hemodiafiltration carried out via different venous catheters, recording a total of 78 measurements. The mean Swan-Ganz cardiac output was 2.22 +/- 0.94 L/min, and mean PiCCO cardiac output was 1.94 +/- 0.80 L/min (no significant difference). The mean difference (bias) of differences (limits of agreement) was 0.2812. The differences between the methods increased with higher cardiac output, but the percentage differences in relation to cardiac output remained stable. Good correlation was found between the two methods: single-measure intraclass correlation was 0.8892 (95% confidence interval, 0.54-0.95). There were no differences between the 5.5- and 7.5-FR Swan-Ganz catheters or between the 4- and 5-Fr PiCCO catheters. Femoral arterial thermodilution cardiac output measurements correlated well with pulmonary arterial thermodilution cardiac output measurements in a pediatric animal model.