The use of assumed values of oxygen consumption has become an accepted practice in the calculation of direct Fick cardiac output. A survey showed that the assumed values in common use were derived from basal metabolic rate studies on normal subjects, a use which may not be valid. We have compared previous assumed values based on basal metabolic rate or cardiac catheterization studies with those obtained by direct measurement in 80 patients (age range 38-78 years) with various cardiac disorders. Comparison of the assumed and directly measured values of indexed oxygen consumption and the cardiac index showed large discrepancies, with over half the values differing by more than +/- 10% and many by more than +/- 25% from the measured value. Assumed values of oxygen consumption should be used with caution when calculating cardiac output during cardiac catheterization procedures, because large errors can result. The equations of LaFarge and Miettinen gave the closest approximation to the measured data and their use is recommended in preference to values predicted from basal metabolic rate studies.