Objective To determine cardiac output from measurements of CO2 production (co2), and arterial (Sao2) and mixed venous (So2) oxygen saturations, using a modified Fick equation, in which cardiac output = co2/[k (Sao2 – So2)], where k represents a constant. Design A metabolic measurement cart was used to measure co2 and oxygen consumption (o2) at 3-min intervals. Sao2 and So2 were measured via a pulse oximeter and a fiberoptic right heart catheter, respectively. The initial value of k for each study was determined from initial simultaneous measurements of thermodilution cardiac output, co2, Sao2, and So2 via the equation k = co2/[cardiac output (Sao2 – Svo2)]. The value of k was assumed to remain constant for the entire study period. Thereafter, cardiac outputs calculated from k and the measurements of co2, Sao2, and So2 were compared with the simultaneously obtained cardiac outputs determined by thermodilution. Similarly, cardiac outputs calculated from the traditional oxygen Fick equation, where cardiac output = o2/[13.4 × hemoglobin (Sao2 – So2)], were compared with the simultaneously acquired cardiac outputs determined by thermodilution. Setting Surgical ICU in a Veterans Affairs Medical Center. Patients Seven postoperative patients, mechanically ventilated using the intermittent mandatory ventilation mode, were studied over a mean period of 4 hrs. Results Cardiac output (obtained from co2 and oximetry saturations) was closely related to thermodilution cardiac output: with linear regression showing r2 = .96 and standard error of the estimate = 0.59 L/min, n = 21; and, with bias and precision = 0.17 and 0.68 L/min, respectively. The traditional oxygen Fick cardiac output was also closely related to the thermodilution cardiac output (r2 = .81, standard error of the estimate = 1.46 L/min, n = 22; bias and precision = 0.31 and 1.46 L/min, respectively). Conclusion The proposed method for calculating cardiac outputs solely from co2 and oximetry saturations yields results that correspond closely to thermodilution determined cardiac outputs. The method is simple and avoids the difficulties in the Fick method associated with accurate o2 measurement. This approach may be suitable for continuous cardiac output monitoring in critically ill patients.