Purpose: the purpose of this study was to evaluate the safety and usability of the ATLAS 2030 in children with Cerebral Palsy (CP) and Spinal Muscular Atrophy (SMA). Materials and Methods: the sample consisted of six children, three with CP and three with SMA, who received eight sessions of robot-assisted gait therapy. Safety was measured by the presence of adverse events. Usability was measured by spatiotemporal parameters, the Six-Minute Walking Test (6MWT), and the time needed for donning and doffing, as well as satisfaction questionnaires administered to therapists and patients. Results: no serious adverse events were reported. The average cadence and number of steps per session increased throughout sessions, as well as the distance covered in the 6MWT, both in participants with CP and SMA. The mean donning time at the end of the study was 4.6 ± 1.3 min, and only one therapist was necessary to carry out all of the sessions. Satisfaction was considered high by both children and therapists. Conclusions: the ATLAS 2030 was shown to be safe for children with CP and SMA. The usability of the device was good, since a progression in the spatiotemporal parameters was observed throughout the sessions, and patient and therapist satisfaction were high.