Non-invasive cardiac output monitoring techniques provide high yield, low risk mechanisms to identify and individually treat shock in the emergency setting. The non-invasive ultrasonic cardiac output monitoring (USCOM) device uses an ultrasound probe applied externally to the chest; however limitations exist with previous validation strategies. This study presents the in vitro validation of the USCOM device against calibrated flow sensors and compares user variability in simulated healthy and septic conditions. A validated mock circulation loop was used to simulate each condition with a range of cardiac outputs (2–10 l/min) and heart rates (50–95 bpm). Three users with varying degrees of experience using the USCOM device measured cardiac output and heart rate by placing the ultrasound probe on the mock aorta. Users were blinded to the condition, heart rate and cardiac output which were randomly generated. Results were reported as linear regression slope (β). All users estimated heart rate in both conditions with reasonable accuracy (β = 0.86–1.01), while cardiac output in the sepsis condition was estimated with great precision (β = 1.03–1.04). Users generally overestimated the cardiac output in the healthy simulation (β = 1.07–1.26) and reported greater difficulty estimating reduced cardiac output compared with higher values. Although there was some variability between users, particularly in the healthy condition (P < 0.01), all estimations were within a clinically acceptable range. In this study the USCOM provided a suitable measurement of cardiac output and heart rate when compared with our in vitro system. It is a promising technique to assist with the identification and treatment of shock.