Adaptation to prismatic displacement was examined under a number of conditions with locomotion in a hallway exposure. In general, total prism adaptation was inversely related to secondary cognitive load (presence or absence of mental arithmetic) and the relative magnitude of visual and proprioceptive shift depended upon the availability of visible sound sources in the hall environment. When the speaking experimenter was visible to the subject, visual shift was greater than proprioceptive shift, but when the experimenter was not visible and]or was silent, proprioceptive shift was greater than visual shift. No relationship has been detected between locomotion (walking rate) and either prism adaptation or cognitive load. These results are consistent with a model which assumes that the total adaptive response depends upon the available capacity to establish and maintain coordinative linkage between discordant systems, whereas locus of adaptive recalibration depends upon the direction of these coordinative linkages, discordance and adaptation occurring in the guided system(s). The available capacity for coordinative linkage depends upon cognitive load, whereas the directionality of linkage is independently determined by task structure. Locomotion is assumed to be mediated by a sensorimotor system that does not involve the distorted positional information.