The objective of this research was to identify variables (demographic, device, neurological, clinical, and training session dose) that were associated with results of the ten-minute walk test (10MWT) for individuals with a spinal cord injury (SCI) who participated in a large randomized crossover clinical trial of exoskeletal walking. Fifty individuals were randomized into Group AB or BA (A = exoskeleton intervention arm, B = control arm). A generalized linear mixed model was applied to model 10MWT and found that a training dose of 36 sessions and gender were the most significant. These variables were more significant than neurological level of injury or completeness of injury. Understanding the effects of exoskeleton/human interface for different devices is crucial for identifying suitable candidates to use the device and developing effective/efficient clinical training protocols for community ambulation, rehabilitation, and recovery post-SCI.