Abstract-The objectives in this study were to investigate the effect of robot-assisted gait training on cardiorespiratory fitness in subjects with motor incomplete spinal cord injury and document the exercise intensity of robotic walking in comparison with the recommended guidelines. Ten patients followed a 24-session training program with a robotic gait orthosis in addition to physiotherapy sessions completed within 10 to 16 wk. Cardiorespiratory fitness was determined in a graded arm crank exercise test before and after the training program. To assess the intensity of robot-assisted walking, oxygen consumption (VO 2 ) and heart rate (HR) were measured during a training session early in and at the end of the training program, and exercise intensity measures (percentage of VO 2 reserve [%VO 2 R], percentage of HR reserve [%HRR], and metabolic equivalents [METs]) were calculated. Whereas no changes were found in peak VO 2 , the resting and submaximal HR at a constant work load were significantly lower after training. Most subjects exercised at low intensity (<30%VO 2 R, <30%HRR, <3.0 METs), and only two subjects exercised at moderate intensity (>3.0 METs). In spite of the low exercise intensity of the training program and no changes in peak VO 2 , robot-assisted gait training induced some improvement in cardiorespiratory fitness, as suggested by lower resting and submaximal HR values.Clinical Trial Registration: ISRCTN Register, ISRCTN67827069, "Recovery of Walking Ability Using a Robotic Device"; http://www. controlled-trials.com/ISRCTN67827069