The aim of the present study was to assess the impact of standardised computer-manipulated changes to images of the orbital region. Judges (n=250) ranked each image for attractiveness on a visual-analogue-scale. The images were thereby classified into different clusters according to their ratings of attractiveness. These clusters were highly significantly different (p<0.001). The degree of negative impact of changes to the orbital region on facial attractiveness increased significantly from lateral to medial and from cranial to caudal. Small asymmetrical or symmetrical lateral variations did not influence attractiveness negatively.