Study design: Each participant completed two submaximal, perceptually guided arm crank exercise tests and a graded exercise test (GXT) to volitional exhaustion. Objective: To assess the validity of a submaximal, perceptually guided exercise test to predict peak oxygen uptake ( . VO 2 peak) during arm cranking in paraplegic individuals. Setting: University of Jordan, Amman, Jordan. Participants: Eleven men with paraplegia as a result of poliomyelitis infection or spinal cord injury completed two submaximal perceptually guided exercise tests and an arm crank GXT to volitional exhaustion. Main outcome measures: The prediction of .VO 2 peak was calculated by extrapolating the submaximal rating of perceived exertion (RPE) and .VO 2 values by linear regression to RPE20. Results: There were no significant differences between measured and predicted .VO 2 peak from the three submaximal ranges of the RPE (that is, 9-13, 9-15 and 9-17) when extrapolated to RPE20 during both perceptually guided exercise tests (all P40.05). However, the second perceptually guided exercise tests provided a more accurate prediction of .VO 2 peak as reflected by narrower 95% limits of agreement and higher intraclass correlation coefficients. Conclusion: This study has shown that .VO 2 peak may be predicted with reasonable accuracy from a perceptually guided exercise test, especially after a full familiarization trial.