Previous investigators demonstrated divergent results, when comparing impedance to other established methods for the assessment of cardiac output. Cardiac output is defined as stroke volume multiplied by heart rate. Heart rate is easy to measure and the main determinant of cardiac output during exercise under physiologic conditions. Therefore, cardiac output (CO) by tetrapolar impedance cardiography (TIC) was compared to thermodilution (TD) in 35 patients (P). In a second step it was examined how heart rate alone would perform as a predictor of CO under exercise. Measurements were done at rest, 25, 50, 75, and 100 w. Results were corrected for hematocrit (HCT). There was no correlation between CO by TD and by TIC at rest (r = -0.006). Under exercise correlation was somewhat better with r = 0.45 despite correction for HCT. Mean values were similar, but scatter was wide. Relative increase in CO by both methods was also correlated (r = 0.46), while heart rate alone correlated best to TD (r = 0.65). Transthoracic impedance cardiography is not a reliable technique to measure absolute values of cardiac output at rest. During exercise large scatter limits this method to the measurement of CO in larger groups. At exercise heart rate alone appears to be a better indicator of increase of cardiac output than impedance cardiography.