This thesis investigated the effect of prior experience on the perception of faces. Firstly, I introduce the methodology used to create an iso-expressive set of face images. These iso-expressive faces were then used to measure sensitivity to facial expression perception in two main experiments in this thesis. One experiment measured facial expression sensitivity to faces that were previously presented as targets or distractors in an attentional task. Then, I present an instrumental learning task used to imbue neutral faces with value. This task was a choice game where participants have to choose the optimal choices in order to maximize payoffs. This technique was used to provide prior experience with faces. I then measured sensitivity to facial expression perception with faces previously associated with those values. Evidence obtc1ined across these experiments supports the notion that facial exp~8ssio!:~ are malleable. Faces associated with rewa1·ds were seen as rrime expressive than f3.ces associated with punishment. This provides evidene,e of an influence of top-down signal in the perception of faces. Top-down signals provided by the value-associated faces were further explored in a visual attention task. Target detection was biased in favour of the motivational salience associated to faces.