This study examines the effect of fascicle length change on motor unit recruitment and discharge rate in the human tibialis anterior (TA) during isometric contractions of various intensities. The torque produced during dorsiflexion and the surface and intramuscular EMGs from the TA were recorded in 8 subjects. The behavior of the same motor unit (n=59) was compared at two ankle joint angles (+10° and -10° around the ankle neutral position). Muscle fascicle length of the TA was measured non-invasively using ultrasonography recordings. When the ankle angle was moved from 10° plantarflexion to 10° dorsiflexion, the torque produced during maximal voluntary contraction (MVC) was significantly reduced (mean ± SD: 35.2 ± 3.3 vs. 44.3 ± 4.2 Nm; P<0.001) and the average surface EMG increased (0.47 ± 0.08 vs. 0.43 ± 0.06 mV; P<0.05). At reduced ankle joint angle, muscle fascicle length declined by 12.7 % (P<0.01) at rest and by 18.9 % (P<0.001) during MVC. Motor units were activated at a lower recruitment threshold for short compared with long muscle fascicle length, either when expressed in absolute values (2.1 ± 2.5 vs. 3.6 ± 3.7 Nm; P<0.001) or relative to their respective MVC (5.2 ± 6.1 vs. 8.8 ± 9.0%). Higher discharge rate and additional motor unit recruitment were observed at a given absolute or relative torque when muscle fascicles were shortened. However, the data indicate that increased rate coding was mainly present at low torque level (<10% MVC), when the muscle-tendon complex was compliant, whereas recruitment of additional motor units played a dominant role at higher torque level and decreased compliance (10-35% MVC). Taken together, the results suggest that the central command is modulated by the afferent proprioceptive information during submaximal contractions performed at different muscle fascicle lengths.