The aim of this study was to investigate the relative contribution of human muscle reoxygenation on force recovery following a maximal voluntary contraction (MVC). Ten athletes (22·9 ± 4·0 years) executed a plantar-flexion sequence including two repeated MVCs [i.e. a 30-s MVC (MVC(30)) followed by a 10-s MVC (MVC(10))] separated by 10, 30, 60, 120 or 300 s of passive recovery. A 10-min passive recovery period was allowed between each MVC sequence. This procedure was randomly repeated with two different recovery conditions: without (CON) or with (OCC) arterial occlusion of the medial gastrocnemius. During OCC, the occlusion was maintained from the end of MVC(30) to the end of MVC(10). Muscle oxygenation (Near-infrared spectroscopy, NIRS, [Hb(diff) ]) was continuously measured during all MVC sequences and expressed as a percentage of the maximal changes in optical density observed during MVC(30). Maximal Torque was analysed at the start of each contraction. Torque during each MVC(10) was expressed as a percentage of the Torque during the previous MVC(30). Torque recovery was complete within 300 s after MVC(30) during CON (MVC(10) = 101·8 ± 5·0%); 88·6 ± 8·9% of the Torque was recovered during OCC (P = 0·005). There was also a moderate correlation between absolute level of muscle oxygenation and Torque (r = 0·32 (90% CI, 0·09;0·52), P = 0·02). Present findings confirm the role of human muscle oxygenation in muscular force recovery during repeated-maximal efforts. However, the correlation between absolute muscle oxygenation and force level during recovery is only moderate, suggesting that other mechanisms are likely involved in the force recovery process.