The purpose of this study was to investigate adaptations in gross mechanical efficiency during asynchronous hand-rim wheelchair propulsion of novice able-bodied participants following 4 weeks of practice. Twenty seven male participants performed a series of five, 4-minute sub-maximal exercise bouts at 1.7 m·s^-1. Arm frequencies consisted of the freely-chosen frequency (FCF), followed by 4 counter-balanced paced trials pushing at 60, 80, 120, and 140% of the FCF. Gross efficiency (GE) was determined. Participants were divided into two experimental groups (FCF, N = 9; 80% FCF, N = 8) and a control group (N = 8). The experimental groups received a 4-week propulsion practice period (3·wk^-1, 12 practice trials) at 1.7 m·s^-1. Post practice period all groups repeated the five 4-minute sub-maximal exercise bouts. Over the practice period the mean GE for the FCF condition increased in both experimental groups (+1.0 & 0.9%) compared to the control group (+0.1%) (P = 0.001). Arm frequency decreased at FCF in both experimental groups (P = 0.001), however, larger changes were observed in the FCF experimental group. Four weeks practice had a beneficial effect on metabolic cost and GE in male novice participants. This improved GE associated to the resulting changes in the self-selected arm frequency in both experimental groups, as an overall indicator of propulsion technique.