The aim of the study was to investigate the adaptations of motor unit discharge rate and additional motor unit recruitment in different regions of the upper trapezius muscle during sustained contraction. Intramuscular EMG signals were recorded from three locations (cranial, middle, and caudal) within the upper trapezius of eleven healthy men during 60-s static shoulder abduction at 25% of the maximal force. Surface EMG signals were recorded concurrently with a 10 × 5 electrode grid. Fifty-one (cranial location), 39 (middle), and 19 (caudal) motor units were identified from the intramuscular EMG signals. Motor unit discharge rates at the beginning of the contraction were greater in the caudal than in the other two locations (cranial: 16.5 ± 3.4 pps; n = 28, middle: 16.2 ± 3.4 pps; n = 32, caudal: 19.6 ± 3.1 pps; n = 13; P < 0.05). Because the decrease in discharge rate over time was larger for the caudal location, at the end of the contraction the discharge rates were comparable at all locations (cranial: 15.7 ± 3.3 pps; n = 28, middle: 14.9 ± 2.3 pps; n = 32, caudal: 15.8 ± 3.6 pps; n = 13). Additional motor unit recruitment was observed in all locations but more frequently in the caudal region (19 motor units recruited after the beginning of the contraction vs. only 4 in each of the other two locations). The mean position of the distribution of surface EMG amplitude over the grid moved by 1.4 ± 0.7 mm (P < 0.001) in the cranial direction at the end with respect to the beginning of the contraction. The results showed that the neural drive received by the upper trapezius depends on the muscle region, which may indicate non-uniform descending drive to the motor neuron pool.