The properties of adaptation within the locomotor and balance control systems directed towards improving one's recovery strategy for fall prevention are not well understood. The purpose of this study was to examine adaptive control of gait stability to repeated slip exposure leading to a reduction in backward loss of balance (and hence in protective stepping). Fourteen young subjects experienced a block of slips during walking. Pre- and post-slip onset stability for all slip trials was obtained as the shortest distance at touchdown (slipping limb) and lift-off (contralateral limb), respectively, between the measured center of mass (COM) state, that is, position and velocity relative to base of support (BOS) and the mathematically predicted threshold for backward loss of balance. An improvement in pre- and post-slip onset stability correlated with a decrease in the incidence of balance loss from 100% (first slip) to 0% (fifth slip). While improvements in pre-slip stability were affected by a proactive anterior shift in COM position, the significantly greater post-slip onset improvements resulted from reductions in BOS perturbation intensity. Such reactive changes in BOS perturbation intensity resulted from a reduction in the demand on post-slip onset braking impulse, which was nonetheless influenced by the proactive adjustments in posture and gait pattern (e.g., the COM position, step length, flat foot landing and increased knee flexion) prior to slip onset. These findings were indicative of the maturing process of the adaptive control. This was characterized by a shift from a reliance on feedback control for postural correction to being influenced by feedforward control, which improved pre-slip stability and altered perturbation intensity, leading to skateover or walkover (>0.05 m or <0.05 m displacement, respectively) adaptive strategies. Finally, the stability at contralateral limb lift-off was highly predictive of balance loss occurrence and its subsequent rapid reduction, supporting the notion of the internal representations of stability limits that could be modified and updated, as a key component in the adaptive control.