Manual wheelchair users exhibit pain related to repetitive and demanding shoulder activities of daily living. Wearable sensing systems like force measurement gloves can provide insights about upper extremity loading in the community and home environments. We calibrated and evaluated the accuracy of a novel force measurement glove with a body-worn data logger. The device was calibrated with loads of 0-800N applied to the palmar surface of the glove. Calibration conditions were tested that varied the stiffness of the material in the glove, the temperature, and the curvature of the force applicator. Calibration equations from each condition were evaluated by comparing the glove's force prediction with the output of an instrumented wheelchair rim during propulsion and weight relief exercises. The force measurement glove detected 72.7% of 355 propulsion peaks and had a strong linear correlation with the instrumented rim force measurements (r=0.80). The most accurate calibration equation was constructed using data from all conditions, with an RMS force measurement error of 64.7 N and 31.7 N for weight relief exercise and propulsion, respectively. The force measurement glove design described here may serve as a useful tool for detection of loading events and relative magnitude changes.