In this study we examined the oxygenation trend of the vastus medialis muscle during sustained high-intensity exercise. Ten cyclists performed an incremental cycle ergometer test to voluntary exhaustion [mean (SD) maximum oxygen uptake 4.29 (0.63) l x min(-1); relative to body mass 60.8 (2.4) ml x kg(-1)min(-1)] and a simulated 20-km time trial (20TT) on a wind-loaded roller system using their own bicycle (group time = 23-31 min) in two separate sessions. Cardiorespiratory responses were monitored using an automated metabolic cart and a wireless heart rate monitor. Tissue absorbency, which was used as an index of muscle oxygenation, was recorded simultaneously from the vastus medialis using near-infrared spectroscopy. Group mean values for oxygen uptake, ventilation, heart rate, respiratory exchange ratio, power output, and rating of perceived exhaustion were significantly (P < or = 0.05) higher during the incremental test compared to the 20TT [4.29 (0.63) l x min(-1) vs 4.01 (0.55) l x min(-1), 120.4 (26) l x min(-1) vs 97.6 (16.1) l x min(-1), 195 (8) beats x min(-1) vs 177 (9) beats min(-1), 1.15 (0.06) vs 0.93 (0.06), 330.1 (31) W vs 307.2 (24.5) W, and 19 (1.5) vs 16 (1.7), respectively]. Oxygen uptake and heart rate during the 20TT corresponded to 93.5% and 90.7%, respectively, of the maximal values observed during the incremental test. Comparison of the muscle oxygenation trends between the two tests indicated a significantly greater degree of deoxygenation during the 20TT [-699 (250) mV vs 439 (273) mV; P < or = 0.05] and a significant delay in the recovery oxygenation from the 20TT. The mismatching of whole-body oxygen uptake and localised tissue oxygenation between the two tests could be due to differences in muscle temperature, pH, localised blood flow and motor unit recruitment patterns between the two tests.