The purpose of this study was to investigate whether athletic and nonathletic manual wheelchair users (MWU) display differences in kinetic and kinematic variables during daily wheelchair propulsion. Thirty-nine manual wheelchair users (athletic n = 25; nonathletic n = 14) propelled their own daily living wheelchair on a roller ergometer at two submaximal speeds for three minutes (1.11 m s−1 and 1.67 m s−1). A 10 camera Vicon motion capture system (Vicon, Motion Systems Ltd. Oxford, United Kingdom) collected three-dimensional kinematics of the upper limbs and thorax at 200 Hz during the final minute of each propulsion trial. Kinetics, kinematics and kinematic variability were compared between athletic and nonathletic groups. Kinematic differences were investigated using statistical parametric mapping. Athletic MWU performed significantly greater physical activity per week compared to nonathletic MWU (920 ± 601 mins vs 380 ± 147 mins, respectively). However, no significant biomechanical differences between athletic and nonathletic MWU were observed during either propulsion speed. During the 1.11 m s−1 trial wheelchair users displayed a stroke frequency of 53 ± 12 pushes/min and a contact angle of 92.5 ± 16.2°. During the 1.67 m s−1 trial the mean stroke frequency was 64 ± 22 pushes/min and contact angle was 85.4 ± 13.6°. Despite the hand being unconstrained during the recovery phase the magnitude of joint kinematic variability was similar across both glenohumeral and scapulothoracic joints during recovery and push phases. To conclude, although athletic MWU participate in more physical activity per week they adopt similar strategies to propel their daily living wheelchair. Investigations of shoulder pain and dailywheelchair propulsion do not need to distinguish between athletic and nonathletic MWU.