Recently, researchers have suggested that when we see an object we automatically represent how that object affords action (Tucker & Ellis, 2001). However, the precise nature of this representation remains unclear: is it a specific motor plan or a more abstract response code? Furthermore, do action representations actually influence what we perceive? In Experiment 1, participants responded to an image of an object and then made a laterality judgment about an image of a hand. Hand identification was fastest when the hand corresponded to both the orientation and grasp type of the object, suggesting that affordances are represented as specific action plans. In Experiment 2, participants saw an image of a hand before interpreting an ambiguous object drawing. Responses were biased towards the interpretation that was congruent with the grasp type of the hand prime. Together, these results suggest that action representations play a critical role in object perception.