The commercially available Perception Neuron motion capture (Mo-Cap) system is a cost effective and easy to use option for motion analysis. However, the accuracy of this system in a practical setting is unknown and needs to be evaluated if it is to be considered for applications that require a specific level of measurement precision. Therefore, the validity of the Mo-Cap system for estimating postural angular kinematics of the upper body was assessed. Upper body motion was evaluated through three-dimensional analysis of functional movements performed by the neck, thorax and shoulders. Range of motion (RoM) estimates were compared to Vicon using Bland-Altman analysis. Systematic biases in neutral to peak RoM differences were all ≤4.5° and random biases ≤±4.5° except for neck extension where the values were larger. The present findings suggest that the Mo-Cap system is a valid method for assessing the majority of upper body ROM to within 5°.