Near-infrared spectroscopy (NIRS) has been used for noninvasive assessment of oxygenation in living tissue. For muscle measurements by NIRS, the measurement sensitivity to muscle () is strongly influenced by fat thickness (FT). In this study, we investigated the influence of FT and developed a correction curve for with an optode distance (3 cm) sufficiently large to probe the muscle. First, we measured the hemoglobin concentration in the forearm () and thigh () during arterial occlusion using a time-resolved spectroscopy (TRS) system, and then FT was measured by ultrasound. The correction curve was derived from the ratio of partial mean optical path length of the muscle layer to observed mean optical path length . There was good correlation between FT and at rest, and could be used to estimate FT. The estimated FT was used to validate the correction curve by measuring the forearm blood flow (FBF) by strain-gauge plethysmography (SGP_FBF) and TRS (TRS_FBF) simultaneously during a reactive hyperemia test with 16 volunteers. The corrected TRS_FBF results were similar to the SGP_FBF results. This is a simple method for sensitivity correction that does not require use of ultrasound.