Understanding fluvial bedload transport and its geomorphic implications is hampered by the paucity and low accuracy of field-derived transport data. Tunnicliffe et al. (Hydrological Processes, 2000; 14: 2631-2643 described the Bedload Movement Detector (BMD), a magnetic induction system for measuring bedload movement in gravel-bed rivers based on a sensor that produces signals when its magnetic field is enhanced by a passing particle. We conducted two types of laboratory experiments to attempt to calibrate the BMD system: (1) rotating platter experiments were designed to relate sensor response curves to particle properties, and (2) flume experiments were designed to measure particle speed and permit computation of a calibrated model. Physically based relations were derived amongst particle volume, speed and magnetic content, and signal integral, amplitude and width. The calibrated relations demonstrate a basis to estimate sediment flux in gravel-bed rivers, but they require independent knowledge of particle magnetic susceptibility. Over the course of the experiments a number of weaknesses in the sensor design and performance were identified and suggestions are made to improve the system. The high spatial and temporal resolution of data that magnetic sensors of bedload movement are likely to offer makes them valuable to consider for sediment transport measurements in both the field and laboratory.