We previously compared the effects of increased respiratory muscle work during whole body exercise and at rest on diaphragmatic fatigue and showed that the amount of diaphragmatic force output required to cause fatigue was reduced significantly during exercise (Babcock et al., J Appl Physiol 78: 1710, 1995). In this study, we use positive-pressure proportional assist ventilation (PAV) to unload the respiratory muscles during exercise to determine the effects of respiratory muscle work, per se, on exercise-induced diaphragmatic fatigue. After 8-13 min of exercise to exhaustion under control conditions at 80-85% maximal oxygen consumption, bilateral phrenic nerve stimulation using single-twitch stimuli (1 Hz) and paired stimuli (10-100 Hz) showed that diaphragmatic pressure was reduced by 20-30% for up to 60 min after exercise. Usage of PAV during heavy exercise reduced the work of breathing by 40-50% and oxygen consumption by 10-15% below control. PAV prevented exercise-induced diaphragmatic fatigue as determined by bilateral phrenic nerve stimulation at all frequencies and times postexercise. Our study has confirmed that high- and low-frequency diaphragmatic fatigue result from heavy-intensity whole body exercise to exhaustion; furthermore, the data show that the workload endured by the respiratory muscles is a critical determinant of this exercise-induced diaphragmatic fatigue.