Living with limb loss comes with numerous physical and psychological ailments, especially in older individuals. Prostheses are commonly used in attempt to alleviate some of these difficulties, but the change in gait compared to age-matched able-bodied individuals and an individual’s ability to learn to use these devices are not clearly understood. This study focuses on older individuals with limb loss and prostheses and evaluates changes in their whole-body and segmental angular momentum (H) as well as center of mass velocity in response to repeated exposures to treadmill-induced gait perturbations. We hypothesized that able-bodied older individuals and older individuals with limb loss using prostheses would both show signs of motor learning, with reduced magnitudes of H and center of mass velocity of successive trials. We also hypothesized that the able-bodied group would perform better, i.e. have lower peak magnitudes of H and center of mass velocity when compared to the limb loss group. Motion capture data was analyzed in Python using Pearson correlation coefficient analysis to evaluate learning over successive trials. The findings confirm the hypotheses. Evidence of learning was found in both groups, with the control group having lower peak magnitudes of H and center of mass velocity. Each individual showed different amounts of learning, with confidence and daily activity level likely influencing one’s ability to learn. The data also showed some individuals with magnitudes trending away from 0, showing an alternative way of learning. These findings have clinical applications and give insight to other research about gait and balance.