We have previously shown that respiratory factors (arterial hypoxemia and/or pulmonary mechanics) contribute to limit maximal incremental exercise in interstitial lung disease (ILD). In this study, we tested the hypothesis that arterial hypoxemia, not pulmonary mechanics, primarily limits maximal exercise in subjects with ILD. Seven subjects with ILD underwent two incremental exercise tests in random order. Test 1: breathing room air (RA); Test 2: breathing 60% O2 with added external dead space (O2VD). Added VD was used to prevent the fall in minute ventilation (VI) while breathing O2. All subjects demonstrated impaired exercise performance (maximal oxygen uptake [VO2], 56 +/- 13% predicted) while breathing RA. There was a significant increase in peak VI (RA, 64.9 +/- 22.3 L/min versus O2VD, 71.0 +/- 20.6; p < 0.05), maximal work rate (RA, 99 +/- 12 watts versus O2VD, 109 +/- 15 watts; p < 0.01), exercise duration (RA, 383 +/- 67 s versus O2VD; 426 +/- 72 s; p < 0.0005) and maximal VO2 (RA, 1.25 +/- 0.21 L/min versus O2VD, 1.39 +/- 0.26; p < 0.05) during the O2VD exercise test. There was a significant correlation between the percent increase in exercise duration during the O2VD test and the DLCO (r = -0.813, p < 0.05). At matched levels of ventilation, subjects demonstrated a significantly deeper and slower pattern of breathing during the O2VD test. Because subjects with ILD were able to further improve their exercise and further increase their VI during the O2VD exercise study, we conclude that arterial hypoxemia, and not respiratory mechanics, predominantly limits maximal incremental exercise in subjects with ILD.