Twenty-four individuals with transtibial amputation were recruited to a randomized, crossover design study to examine stride-to-stride fluctuations of lower limb joint flexion/extension time series using the largest Lyapunov exponent (λ). Each individual wore a "more appropriate" and a "less appropriate" prosthesis design based on the subject's previous functional classification for a three week adaptation period. Results showed decreased λ for the sound ankle compared to the prosthetic ankle (F1,23 = 13.897, p = 0.001) and a decreased λ for the "more appropriate" prosthesis (F1,23 = 4.849, p = 0.038). There was also a significant effect for the time point in the adaptation period (F2,46 = 3.164, p = 0.050). Through the adaptation period, a freezing and subsequent freeing of dynamic degrees of freedom was seen as the λ at the ankle decreased at the midpoint of the adaptation period compared to the initial prosthesis fitting (p = 0.032), but then increased at the end compared to the midpoint (p = 0.042). No differences were seen between the initial fitting and the end of the adaptation for λ (p = 0.577). It is concluded that the λ may be a feasible clinical tool for measuring prosthesis functionality and adaptation to a new prosthesis is a process through which the motor control develops mastery of redundant degrees of freedom present in the system.