Arterial desaturation impairs exercise performance in a dose-dependent manner. However, new theories of exercise-induced fatigue suggest that increasing rates of arterial deoxygenation augment the fatigue response during exercise. The purpose of this dissertation is to clarify if self-selected exercise intensity, while exercising at a constant rate of perceived exertion (RPE), is sensitive to alterations in the absolute arterial saturation (SPO2) and/or the rate of change in SPO2. Subjects performed constant RPE exercise for 30 min. They were instructed to adjust their exercise intensity during the trial to maintain their RPE at 5 on Borg’s 10-point scale. Subjects engaged in continuous bilateral, isokinetic cycling and intermittent, unilateral, isometric knee-extension. The fraction of inspired oxygen was reduced to desaturate arterial blood from starting values (>98%) to 70%. This desaturation occurred linearly over 3 target time periods (FAST, 5 min; MED, 15 min; SLOW, 25 min). The rate of arterial desaturation was significantly different between each of the 3 conditions. During cycling exercise, PO (FAST = 2.8 ± 2.1 W·% SPO2−1; MED = 2.5 ± 1.8 W·% SPO2−1; SLOW = 1.8 ± 1.6 W·% SPO2−1; P < 0.001) and surface electromyography (sEMG) of the vastus medialis (FAST = 1.3 ± 0.6%·% SPO2−1; MED = 1.1 ± 0.5%·% SPO2−1; SLOW = 0.7 ± 0.7%·% SPO2−1; P < 0.001) decreased at significantly different rates. Post hoc comparisons revealed that the rates of decline in PO and sEMG during FAST and MED were similar, and both were greater than SLOW. However, during isometric knee extension exercise, the level of force production and sEMG remained similar across saturation levels. These results confirm that decreases in absolute SPO2 impair self-selected exercise intensity and that faster desaturation rates magnify that impairment, but only when a large muscle mass is engaged. These findings suggest that the rate of arterial deoxygenation independently influences exercise performance and that the central depressant effect may be a function of the metabolic strain associated with hypoxia, rather than the hypoxia per se.