To quantitate the O2 cost of maximal exercise hyperpnea, we required eight healthy adult subjects to mimic, at rest, the important mechanical components of submaximal and maximal exercise hyperpnea. Expired minute ventilation (VE), transpulmonary and transdiaphragmatic (Pdi) pressures, and end-expiratory lung volume (EELV) were measured during exercise at 70 and 100% of maximal O2 uptake. At rest, subjects were given visual feedback of their exercise transpulmonary pressure-tidal volume loop (W(v)), breathing frequency, and EELV, which they mimicked repeatedly for 5 min per trial over several trials, while hypocapnia was prevented. The change in total body O2 uptake (VO2) was measured and presumed to represent the O2 cost of the hyperpnea. In 61 mimicking trials with VE of 115-167 l/min and W(v) of 124-544 J/min, VE, W(v), duty cycle of the breath, and expiratory gastric pressure (Pga) integrated with respect to time (integral Pga.dt/min) were not different from those observed during maximum exercise. Integral Pdi.dt/min was 14% less and EELV was 6% greater during maximum exercise than during mimicking. The O2 cost measurements within a subject were reproducible over 3-12 trials (coefficient of variation +/- 10%, range 5-16%). The O2 costs of hyperpnea correlated highly and positively with VE and W(v) and less, but significantly, with integral Pdi.dt and integral Pga.dt. The O2 cost of VE rose out of proportion to the increasing hyperpnea, so that between 70 and 100% of maximal VO2-DELTA-VO2/DELTA-VE increased 40-60% (1.8 +/- 0.2 to 2.9 +/- 0.1 ml O2/l VE) as VE doubled. We believe that these data provide a realistic conservative estimate of the actual cost of exercise hyperpnea over a broad range of ventilatory outputs during moderately heavy-to-maximum exercise in healthy human subjects.