In this study, we used a model-based approach to explore the potential contributions of central pattern generating circuits (CPGs) during adaptation to external perturbations during locomotion. We constructed a neuromechanical modeled of locomotion using a reduced-phase CPG controller and an inverted pendulum mechanical model. Two different forms of locomotor adaptation were examined in this study: split-belt treadmill adaptation and adaptation to a unilateral, elastic force field. For each simulation, we first examined the effects of phase resetting and varying the model's initial conditions on the resulting adaptation. After evaluating the effect of phase resetting on the adaptation of step length symmetry, we examined the extent to which the results from these simple models could explain previous experimental observations. We found that adaptation of step length symmetry during split-belt treadmill walking could be reproduced using our model, but this model failed to replicate patterns of adaptation observed in response to force field perturbations. Given that spinal animal models can adapt to both of these types of perturbations, our findings suggest that there may be distinct features of pattern generating circuits that mediate each form of adaptation.