Abstract Background Shoulder muscle overuse has been linked to the high prevalence of shoulder injuries in manual wheelchair users. Ramp ascent is a barrier that is often faced by manual wheelchair users that requires higher muscle activations than level wheelchair propulsion. Additionally, reported subjective measures of shoulder pain are reduced amongst manual wheelchair users when using a wheelchair wheel outfitted with a gear mechanism. The purpose of the current investigation was to investigate shoulder muscle activation levels and upper limb kinematics during ramp ascent with and without the use of a geared wheel. Methods Thirteen healthy participants (6 male and 7 female) performed ramp ascent on four ramp grades (1:12, 1:10, 1:8, 1:6) using three wheel conditions (gear, no gear, standard). Electromyographic (EMG) activity of select shoulder muscles as well as kinematics of the right upper limb were collected during ramp ascent. Peak and integrated EMG as well as peak wrist, elbow, and shoulder kinematics were obtained from all ramp ascent trials. Findings Peak EMG of the shoulder flexors decreased by an average of 17% ( P ⩽ 0.0229) during ramp ascent with the geared wheel. Integrated EMG increased by 67% ( P ⩽ 0.0034) as a consequence of an 86% increase ( P = 0.0009) in ramp ascent duration during the geared wheel condition. There were no significant differences between the non-geared and standard wheel conditions. Interpretation Caution must be used if using the gear ratio for prolonged periods due to potential for muscle fatigue since the overall muscle effort to move a fixed distance is higher with the gear ratio. Reducing peak demands may benefit wheelchair users with performing more strenuous tasks of daily living.