Noninvasive measurement of cardiac output (QT) is problematic during heavy exercise. We report a new approach that avoids unpleasant rebreathing and resultant changes in alveolar PO2 or PCO2, by measuring short-term acetylene (C2H2) uptake by an open-circuit technique, with application of mass balance for the calculation of QT. The method assumes that alveolar and arterial C2H2 pressures are the same, and we account for C2H2 recirculation by extrapolating end-tidal C2H2 back to breath I of the maneuver. We correct for incomplete gas mixing by using He in the inspired mixture. The maneuver involves switching the subject to air containing trace amounts of C2H2 and He ventilation and pressures of He, C2H2, and CO2 are measured continuously (the latter by mass spectrometer) for 20-25 breaths. Data from three subjects for whom multiple Fick O-2 measurements of QT were available showed that measurement of QT by the Fick method and by the C2H2 technique was statistically similar from rest to 90% of maximal O-2 consumption (VO2max). Data from 12 active women and 12 elite male athletes at rest and 90% of VO2max fell on a single linear relationship, with O-2 consumption (VO2) predicting QT values of 9.13, 15.9, 22.6, and 29.4 l/min at VO2 Of It 2, 3, and 4 I/min. Mixed venous Po, predicted from C2H2-determined QT, measured VO2 and arterial O-2 concentration was similar to 20-25 Torr at 90% of VO2max during air breathing and 10-15 Torr during 13% O-2 breathing. This modification of previous gas up take methods, to avoid rebreathing, produces reasonable data from rest to heavy exercise in normal subjects.