The hand-wrist region is reported as the most common injury site in boxing. Boxers are at risk due to the amount of wrist motions when impacting training equipment or their opponents, yet we know relatively little about these motions. This paper describes a new method for quantifying wrist motion in boxing using an electromagnetic tracking system. Surrogate testing procedure utilising a polyamide hand and forearm shape, and in vivo testing procedure utilising 29 elite boxers, were used to assess the accuracy and repeatability of the system. 2D kinematic analysis was used to calculate wrist angles using photogrammetry, whilst the data from the electromagnetic tracking system was processed with visual 3D software. The electromagnetic tracking system agreed with the video-based system (paired t tests) in both the surrogate (< 0.2°) and quasistatic testing (< 6°). Both systems showed a good intraclass coefficient of reliability (ICCs > 0.9). In the punch testing, for both repeated jab and hook shots, the electromagnetic tracking system showed good reliability (ICCs > 0.8) and substantial reliability (ICCs > 0.6) for flexion-extension and radial-ulnar deviation angles, respectively. The results indicate that wrist kinematics during punching activities can be measured using an electromagnetic tracking system.