Thoracic electrical bioimpedance (TEB) is a noninvasive method for the estimation of left ventricular stroke volume (SV). Objections to TEB in the past have included systematic overestimation of SV and cardiac output in normals, poor correlation in absolute terms with standard cardiac output reference techniques, motion and ventilation artifacts which distorted recorded waveforms except when patients remained in apnea during measurements, and lack of a computer to yield rapid heat-to-beat quantitation of SV and cardiac output. The present study entails real-time, beat-to-beat cardiac output determinations using TEB and assessing its agreement via a digital computer against a reference standard, thermodilution. A correlation coefficient of 0.88 is reported with 85% of data points falling within 20% confidence limits on either side of the line of identity.