BACKGROUND: The accuracy of currently available devices using pulse contour analysis without external calibration for cardiac index (CI) estimation is negatively impacted by hyperdynamic states, low systemic vascular resistance (SVR), and abrupt changes in SVR. The aim of this study was to evaluate the accuracy of a new device, the Pulsioflex (Pulsion Medical System), in patients undergoing liver transplantation. METHODS: Thirty consecutive patients scheduled for liver transplantation were included. CI was monitored using pulmonary arterial catheter (CI-PAC) and Pulsioflex (CI-Pulsio). Simultaneous CI measurements were made intraoperatively at 9 different stages of the procedure. RESULTS: Two hundred seventy pairs of measurements were analyzed. The median CI-Pulsio values (3.3; interquartile range, 2.8-3.8 L·min·m) were significantly different from the median CI-PAC (4.1; interquartile range, 3.1-5.0 L·min·m; P < .0001). Bland and Altman analysis showed a mean bias of 0.8 L·min·m and 95% limit of agreement from -2.5 to 4.1 L·min·m. Percentage error was 65% (95% confidence interval, 60%-71%). Considering the variations in CI between 2 stages, the comparison between changes in CI-PAC and changes in CI-Pulsio showed a mean bias of 0.1 L·min·m and 95% limit of agreement of -2.1 to 2.2 L·min·m. When excluding changes in CI <0.5 L·min·m (154 paired analyzed), the concordance rate was 62% (95% confidence interval, 54%-70%). The bias between CI-PAC and CI-Pulsio was negatively correlated with SVR (r = -0.67, P < .0001). The bias between changes in CI-PAC and changes in CI-Pulsio was also negatively correlated with changes in SVR (r = -0.52, P < .0001). CONCLUSIONS: In patients undergoing liver transplantation, Pulsioflex does not accurately estimate CI. Its accuracy is highly impacted by SVR, and it is not able to track changes in CI when large variations in SVR occur.