Study design: Cross-sectional study design. Background: Cardiopulmonary exercise testing (CPET) using the upper-body is an alternative to standardised CPET. A variety of test protocols has been developed for CPET performed on an arm crank for individuals with lower limb disabilities. However, no test protocol specifically designed to evaluate exercise capacity in younger adults with lower limb amputees is available. Objective: To determine test–retest reliability of VO2peak, watts, test time and max heart rate for an upper-body CPET. Methods: Males with unilateral lower limb amputation at crus-level or above, aged 18–40 years, without comorbidities preventing arm cycling were eligible. Participants completed three maximal tests with a minimum of two days between tests and maximum 14 days between the first and third test. Results: Eight participants volunteered to participate and were tested three times. Excellent and acceptable ICC values were found for power in Watts (ICC = 0.73) and test time (ICC = 0.74). Good but non-acceptable test–retest reliability was found for VO2peak (ICC = 0.51) and max heart rate (ICC = 0.64) for all three test sessions. Conclusions: Upper-body CPET shows acceptable test–retest reliability for Watts and test time in younger adult lower limb amputees, but non-acceptable reliability for VO2peak and heart rate.