The primary somatosensory area of the mammalian cerebral cortex is known to play a key role in tactile sensory processing and touch perception, which may not be restricted to an efficient analysis of the incoming sensory flow. Indeed, the neocortex is thought to be capable of generating predictions about future events based on the context, our past experiences and current motor actions. According to predictive coding theories, sensory perception would rely on the continuous comparison of expected sensory inputs with those actually received, and the error signal generated in case of mismatch between these two types of information would be key to optimize behavior.With the aim of studying the predictive mechanisms at play in tactile sensory perception using the mouse whisker system as a model, we designed a whisker-guided locomotion task where we can experimentally introduce deviances between expected and received tactile inputs. We collect calcium sensitive signals over the primary somatosensory area through a bundle of optical fibers in mice trained to navigate and avoid obstacles in predictable spatial arrangements, in complete darkness. We analyze cortical dynamics recorded in such familiar contexts, and when an obstacle is suddenly removed, thus creating a mismatch between expected and received tactile inputs. By simultaneously recording the animal’s movements and whisking strategy using high-speed videography, we extract behavioral markers of expectancy and link them with cortical dynamics recorded synchronously, at the millisecond timescale, whilst the animal is navigating and gathering tactile information in an ethologically relevant manner.