Two types of lower extremity exoskeleton rehabilitation training robots (named Aiwalk and Ailegs) had been the first powered gait orthosis (PGO) in China. The aim of this study conducted in patients with paraplegia was to verify the effectiveness and safety of the lower extremity exoskeleton rehabilitation training robot developed in China. Forty subjects (nine females and thirty-one males with the age of 38.1 ± 9.4 years, height of 169.8 ± 6.5 cm, weight of 62.2 ± 7.8 kg and the post-injury time of 67.8 ± 112.8 months) with T6–L2 spinal cord lesions participated in the study. The patients were characterized by a traumatic SCI with scale A (twenty-nine patients), B (four patients) and C (seven patients) according to American Spinal Injury Association score. All patients received a walking and standing program with the isocentric reciprocating gait orthosis (IRGO) and the PGOs. After 1 week of gait training, subjects walked using three types of orthosis in random order at their self-selected speed each half day and repeated ten times. A wireless single channel ECG recorder and a stop watch were used to measure the speed of walking and heart rate. Walking speed, the distance walked, the physiological cost index (PCI), orthosis wearing speed and rating of perceived exertion all improved with both PGOs as compared with the IRGO. Adverse events associated with PGO were calcaneus compression fracture (one patients) and mild skin pressure ulcer (two patients). Patients with paraplegia (below T6 level) were able to ambulate efficiently using the powered exoskeleton for overground ambulation, providing potential use for functional gain and improved fitness. Fracture and skin damage may be the main risk focus in clinical application.