Most research in human locomotion is limited to steady-state and constant-speed conditions. However, moving about in everyday life requires us to constantly adapt our locomotion strategies in response to intrinsic noise-like transients, uncertainties and external environmental irregularities. In this talk, I will discuss our work on understanding and predicting such adaptive locomotion behaviors in three separate studies: I. The energetics of walking while changing speeds, II. control strategies for running stably in the presence of noise-like deviations and III. Energetics and stability of learning to walk asymmetrically. In part I, we find that the metabolic cost of walking with varying speeds is significant and predicts short-distance overground walking behavior in non-amputees and amputees. In part II, we show that relatively small variability during constant-speed running contains information about stability and control. And such a controller, inferred from variability in human data, can control a simple biped model for perturbations ten times larger than the range from which it was derived. And in part III, by combining energy-optimality and stability of a simple biped model, we predict long-term and short-term adaptations, respectively, observed in split-belt walking behavior.