Background: Manual wheelchair propulsion is often associated with pain in the upper extremities. Recording spatio-temporal parameters can optimize movement patterns and prevent injuries. This study compares a marker-based camera system with inertial measurement units to validate their use in wheelchair propulsion on a test stand. Methods: Spatio-temporal parameters of 27 manual wheelchair users propelling at three self-selected speeds (slow, normal, fast) were simultaneously recorded using a marker-based camera system and inertial measurement units, and subsequently compared between both systems. Results: A high correlation was observed among all spatio-temporal parameters (ρ > 0.992). The biases for the start time of hand contact with the pushrim (−0.02 ± 0.02 s), hand release from the pushrim (−0.02 ± 0.01 s), and push length (−0.45 ± 21.45 ms) were slightly overestimated, while recovery length (0.54 ± 21.02 ms), cycle speed (2.37 ± 2.67°/s), and push angle (1.75 ± 4.14°) were slightly underestimated. No bias was found for propulsion frequency. Conclusions: The spatio-temporal parameters recorded using inertial measurement units are suitable for the evaluation of manual wheelchair propulsion and can be used in a clinical context. The low acquisition costs and simple installation process may increase the use of inertial measurement units in the future.