OBJECTIVE To assess the accuracy of thermodilution cardiac output measurements from the right ventricular port vs. the central venous port. In addition, waveform patterns were evaluated in 50 right-heart catheters to determine the actual location of the right ventricular port. METHODS Central venous port cardiac output measurements were compared with right ventricular port cardiac output measurements using the same right-heart catheter. METHODS The general ICU of Memorial Sloan-Kettering Cancer Center. METHODS Forty-seven critically ill cancer patients with 60 different right-heart catheters were evaluated. METHODS Four injections of 10 mL of iced normal saline were made through each port, with the results of the last three injections averaged. Cardiac output determinations from both ports were completed in less than 10 min. The order of port injection was random. RESULTS No difference was noted between cardiac output determinations from the two ports (paired t-test). Twenty-five of 50 right-heart catheters were in the right ventricle, with the other 25 in the right atrium. A comparison of ports in the 25 catheters that were in the right ventricle showed no difference with a significant (p less than .01, r2 = .94) correlation. CONCLUSIONS Thermodilution cardiac output measurements using 10 mL of iced saline can be determined accurately using the right ventricular port if the central venous port becomes nonfunctional.