We predict how our actions will influence the world around us. Prevailing models of action control propose that we use these predictions to suppress or ‘cancel’ perception of expected action outcomes. However, contrasting normative Bayesian models in sensory cognition suggest that top-down predictions bias observers toward perceiving what they expect. In this talk I will describe neuroimaging (fMRI) and psychophysical experiments comparing these two accounts of perceptual prediction. We find that observers are biased toward perceiving expected action outcomes, and that predicted action outcomes are associated with higher fidelity (‘sharper’) representations in visual brain areas. These results are difficult to reconcile with classic ‘cancellation’ theories, and instead reveal a gain control mechanism that can explain how we generate largely veridical representations of our actions and their consequences in an inherently uncertain sensory world.