Markerless motion capture, due to its low cost and ease of use, has potential to be used as a surrogate for existing marker-based motion capture systems, and may be useful as a supplementary tool for existing biomechanical assessments where no technology is currently used. The accuracy of a dual-camera markerless motion capture system, utilizing a pair of Stereolabs ZED 2i cameras, was compared to a marker-based motion capture system. Joint angles over time were computed for both systems, and a Bland-Altman analysis was performed on both data streams. The joint angle accuracy of the markerless motion capture system over a set of simple movements is presented. The RMSE of the markerless system ranged from 6.2°–16.9° for leg and ankle angles, 3.5–20.6 for trunk/pelvic angles, 9.5–31.2 for head angles, and 11.8–29.9 for arms. This indicates that this system is likely unsuitable as a surrogate for existing clinical motion capture systems, and that it is unlikely that this technology is suitable for novel motion capture applications, such as within functional capacity evaluations.