The conventional continuous wave near-infrared spectroscopy (CW-NIRS) has enabled identification of regional differences in muscle deoxygenation following onset of exercise. However, assumptions of constant optical factors (e.g., path length) used to convert the relative changes in CW-NIRS signal intensity to values of relative concentration, bring the validity of such measurements into question. Furthermore, to justify comparisons among sites and subjects, it is essential to correct the amplitude of deoxygenated hemoglobin plus myoglobin [deoxy(Hb+Mb)] for the adipose tissue thickness (ATT). We used two time-resolved NIRS systems to measure the distribution of the optical factors directly, thereby enabling the determination of the absolute concentrations of deoxy(Hb+Mb) simultaneously at the distal and proximal sites within the vastus lateralis (VL) and the rectus femoris muscles. Eight subjects performed cycle exercise transitions from unloaded to heavy work rates (>gas exchange threshold). Following exercise onset, the ATT-corrected amplitudes (A(p)), time delay (TD(p)), and time constant (τ(p)) of the primary component kinetics in muscle deoxy(Hb + Mb) were spatially heterogeneous (intersite coefficient of variation range for the subjects: 10-50 for A(p), 16-58 for TD(p), 14-108% for τ(p)). The absolute and relative amplitudes of the deoxy(Hb+Mb) responses were highly dependent on ATT, both within subjects and between measurement sites. The present results suggest that regional heterogeneity in the magnitude and temporal profile of muscle deoxygenation is a consequence of differential matching of O(2) delivery and O(2) utilization, not an artifact caused by changes in optical properties of the tissue during exercise or variability in the overlying adipose tissue.