Due to the invasiveness of direct techniques and problems that constrain the use of popular indirect techniques during very heavy (non-steady-state) exercise, measurements of maximal cardiac output are seldom included in studies of exercise. The acetylene-rebreathing technique is well-suited for maximal exercise; however, until recent technological advances, difficulties involved in collecting and measuring alveolar acetylene samples have restricted its use. We compared cardiac output values measured via the acetylene-rebreathing technique (QA) (modified for use with a mass spectrometer) and the dye-dilution technique (QD) at rest and during light to maximal exercise in six moderately active males. Although QA consistently underestimated QD, the two techniques showed a significant correlation of 0.87 throughout all levels of exercise. During maximal exercise, QA and QD values were not significantly different (24.7 and 26.7 l X min-1, respectively). Modifications in the QA technique that reduce potential sources of error were also examined. We conclude that the acetylene-rebreathing technique, modified for use with a mass spectrometer, is a simple and valid procedure for measuring maximal cardiac output.