The study investigated whether the capacity to regulate muscle blood flow (Q) relative to metabolic demand (VO2) is impaired in COPD. Using six NIRS optodes over the upper, middle and lower vastus lateralis in 6 patients, (FEV1:46 ± 12%predicted) we recorded from each: a) Q by indocyanine green dye injection, b) VO2/Q ratios based on fractional tissue O2 saturation and c) VO2 as their product, during constant-load exercise (at 20%, 50% and 80% of peak capacity) in normoxia and hyperoxia (FIO2:1.0). At 50 and 80%, relative dispersion (RD) for Q, but not for VO2, was greater in normoxia (0.67 ± 0.07 and 0.79 ± 0.08, respectively) compared to hyperoxia (0.57 ± 0.12 and 0.72 ± 0.07, respectively). In both conditions, RD for VO2 and Q significantly increased throughout exercise; however, RD of VO2/Q ratio was minimal (normoxia: 0.12–0.08 vs hyperoxia: 0.13–0.09). Muscle Q and VO2 appear closely matched in COPD patients, indicating a minimal impact of heterogeneity on muscle oxygen availability at submaximal levels of exercise.