This paper proposes a complete motion sensing solution for wheelchairs with the main objective of determining tips, falls and other hazardous situations. This kind of motion-logging solution is necessary to study needs and habits of wheelchair users. The system collects motion information utilizing an inertial measurement unit (IMU) consisting of three orthogonal accelerometers and two gyroscopes sensing pitch and roll rate. The IMU is part of a portable device (the “Motion Logger”) which captures raw motion data in a Secure Digital memory card. An algorithm running in the Motion Logger helps reducing energy consumption and memory usage. Actual determination of major events is carried out off-line and in batch mode using a personal computer application, which has two key components: an attitude estimation algorithm and the event identification itself. This document describes the design process, the hardware, and the main algorithms employed in the motion sensing solution, as well as results obtained with it.