Facial attractiveness is often studied on the basis of the internal facial features alone. This study investigated how this exclusion of the external features affects the perception of attractiveness. We studied the effects of two most commonly used methods of exclusion, where the shape of an occluding mask was defined by either the facial outline or an oval. Participants rated attractiveness of the same faces under these conditions. Results showed that faces were consistently rated more attractive when they were masked by an oval shape rather than by their outline (Experiment 1). Attractive faces were more strongly affected by this effect than were less attractive faces when participants were able to control the viewing time. However, unattractive faces benefited more from this effect when the same face stimuli were presented briefly for only 20 ms (Experiment 2). Further manipulation confirmed that the effect was mainly due to the occlusion of a larger area of the external features rather than the regular and symmetrical features of the oval shape (Experiment 3) or lacks contextual cues about the face boundary (Experiment 4). The effect was only relative to masked faces, with no advantage over unmasked faces (Experiment 5), and is likely a result of the interaction between the shape of a mask and the internal features of the face. This holistic effect in the appraisal of facial attractiveness is striking because the oval shape of the mask is not a part of the face but is the edge of an occluding object.