Objective. To study whether values for peak oxygen uptake (VO2peak) and work economy (WE) at a standardized workload are different when tested by arm crank ergometry (ACE) and wheelchair ergometry (WCE). Methods. Twelve paraplegic men with spinal cord injury (SCI) in stable neurological condition participated in this cross-sectional repeated-measures study. We determined VO2peak and peak power output (POpeak) values during ACE and WCE in a work-matched protocol. Work economy was tested at a standardized workload of 30 Watts (W) for both ACE and WCE. Results. There were no significant differences in VO2peak (mL·kg(-1)·min(-1)) between ACE (27.3 ± 3.2) and WCE (27.4 ± 3.8) trials, and a Bland-Altman plot shows that findings are within 95% level of agreement. WE or oxygen consumption at 30 W (VO2-30W) was significantly lower during WCE compared to ACE (P < 0.039). Mean (95% CI) POpeak (W) were 130 (111-138) and 100 (83-110) during ACE and WCE, respectively. Conclusion. The findings in the present study support the use of both ACE and WCE for testing peak oxygen uptake. However, WE differed between the two test modalities, meaning that less total energy is used to perform external work of 30 W during wheelchair exercise when using this WCE (VP100 Handisport ergometer). Clinical Trials Protocol Record is NCT00987155/4.2007.2271.