In this paper, we propose an ultra wide band (UWB)-based full-body motion capture system for gait analysis. The system is designed to reduce the implementation complexity and cost of current gait analysis setups, which require expensive locomotion laboratories and equipment. The solution exploits recent developments in wearable computing (i.e., electronic textiles) and relies on estimating the relative distance between body-worn sensors through time-of-flight measurements. While such a system may find application in several areas, substantial research is still required to prove its feasibility and accuracy. From this perspective, the present contribution addresses two key aspects to enable its development: i) the interaction between UWB electromagnetic pulses and the human body is analyzed and, by contrast to most results in the literature, the body- induced pulse distortion is characterized in terms of an equivalent channel impulse response, and ii) by explicitly taking into account that body-induced distortion, the accuracy of various time-of- flight-based ranging algorithms is investigated.