A common view in philosophy and science, with origins in antiquity, is that we perceive things – see and hear them, for instance – by having representations or images of them, of which we are directly aware. We shall consider the arguments for this view and show that they fail. These include arguments from illusion and hallucination, from the alleged possibility of an inverted spectrum, the Knowledge Argument, claims about what brains in a vat can think and experience, ideas about phantom limb pains, afterimages, and more. Along the way, zombies will be exorcised. We shall also look at arguments coming from physics, such as the time-lag argument. Time permitting, we shall also study Wittgenstein’s private-sensation language argument, criticising the meaningfulness of talk about private representations. Our conclusion will be that no good reason has been provided to assume that perception is via representations.