The face represents an important source of social information in humans, which provides cues to attention, emotion, age, and identity. Past research has shown that sexual dimorphism cues of applicants’ faces and facial attractiveness affect leadership selection: A masculine facial appearance is congruent with masculine-typed leadership roles. Thus, it is extremely significant to examine the face traits about certain character. The purpose of the current study is to investigate the facial image of emperor from the perspective of Confucian culture. In the present study, we chosen a total of 90 participants of college students and adopted 2(Participant Gender: men vs. women)×2(Facial attractiveness: high vs. low) mixed experimental design to inspect the impact of facial attractiveness on the preferences for the face of emperor. Dependent variables were a forced-choice decision for the composite faces and the certainty of decision. These pairs of composite faces were produced by FaceGen Modeller 3.1 and by PhotoShop CS5. Each pair of composite faces was made of two different gender typicality of faces which was produced by the same original facial picture. In the experiment, let participants to choose the faces that are coincident with the images of emperor in their minds. The black fixation was presented firstly on the screen and the time of presentation was 1000ms. Then, certain pair of composite faces was presented on the screen until the participants made their decision. These faces were presented to the participants randomly. For each pair of faces, participants were asked to choose the face which looked more like a emperor. Participants were also instructed to indicate the strength of his decision by choosing from the options ‘slightly more coincident’,‘somewhat more coincident’, ‘more coincident’ and ‘much more coincident’. If the participants recognized the left face as emperor, they were required to choose “1–4”. If the participants recognized the right face as emperor, they were required to choose “5–8”. Trial order and the side of the screen on which any particular image was shown were fully randomized. After viewing and deciding on these portraits the sequence was rerun to rate these faces for masculinity/femininity and warmth. In the experiment, the results revealed that the image of emperor was feminine in one’s mind. (1)When judging these composite faces, one-sample t-tests comparing the strength of coincidence to the emperor’s face in their minds with what would be expected by chance alone(i.e. 3.5) showed that participants judged feminine faces as being more coincident than masculine faces. (2) a 2 × 2 mixed model analysis of variance found a main effect of facial attractiveness: the preference for the emperor’s face was more pronounced for those faces with low attractiveness. Taken together, the present conclusions are as follows:(1) The feminine-looking male face is coincident with the emperor's image in one’s mind. (2) Comparing to the high attractive feminine-looking male face, the low attractive feminine-looking male face is more coincident with the emperor's image in one’s mind.