The purpose of this study was to determine the contributions of arm frequency and propulsion mode on the internal work during submaximal wheelchair propulsion. Twelve able-bodied participants performed a V.O (2) peak test on a wheelchair ergometer. On a separate occasion, six (4 min) submaximal exercise conditions employing two modes of propulsion (synchronous, SYN vs. asynchronous, ASY) at arm frequencies of 40 and 80 rev . min (-1) were performed at 1.2 m . s (-1) and 1.7 m . s (-1). These conditions resulted in three push strategy combinations (ASY [20 : 20], SYN [40 : 40] & ASY [40 : 40]) at two speeds. Gross, net, work and delta efficiency were determined. The cost of unloaded exercise was significantly lower for the ASY [20 : 20] than both ASY and SYN [40 : 40] (0.49 vs. 0.58 and 0.57 L . min (-1), respectively). All the efficiency indices decreased as velocity increased (p < 0.01). ASY [20 : 20] was the least efficient (gross and work) mode (4.2 +/- 0.4 % and 6.2 +/- 0.8 % respectively). Comparison of equal arm frequencies (ASY [40 : 40] vs. SYN [40 : 40]); found the efficiency to be lower for ASY propulsion (p < 0.05). Under the current testing conditions SYN propulsion mode offers greater efficiency during wheelchair propulsion.