This paper presents a wearable inertial-sensing-based body sensor network (BSN) composed of two inertial modules that are placed on human upper limb for real-time human motion capture applications. Each inertial module consists of an ARM-based 32-bit microcontroller (MCU), a triaxial accelerometer, a triaxial gyroscope, and a triaxial magnetometer. To estimate shoulder range of motion (ROM), the accelerations, angular velocities, and magnetic signals are collected and processed by a quaternion-based complementary nonlinear filter for minimizing the cumulative errors caused by the intrinsic noise/drift of the inertial sensors. The proposed BSN is a cost-effective tool and can be used anywhere without any external reference device for shoulder ROM. The sensor fusion algorithm can reduce orientation error effectively and thus can assess shoulder joint motions accurately.