Two methodologies have traditionally been used to measure hemispheric specialization for perception and expression of emotions in human and nonhuman primates. The first refers to objective measures, that is, measures of area and length of facial features, and the second concerns subjective “measures,” that is, assessment of chimeric faces by human judges. We proposed a refined approach to the subjective assessment of hemispheric specializations, which aims at delimiting methodological issues in the study of orofacial asymmetries. The study focused on a baboon threat orofacial expression, which has led to discordant results according to the methodology used (Wallez & Vauclair, 2011, 2013). We presented human participants with two sets of chimeric stimuli varying the regions of the face likely to be processed. The whole face set was composed of classical chimeric faces, and the upper face set was made of chimeric faces for which the lower part was blurred. The purpose of this new procedure was to shed light on the perception process of baboon faces by human participants during a free-viewing chimeric task. Results showed a concomitant influence of the chimera structure depending on the order of the presentation of the set, revealing a training effect in our human judges. These factors combined together allowed the appearance of an overall left–left chimeric choice by human judges (i.e., which indicates a right hemisphere involvement in baboon threat expression). These findings bring novel insights into the study of orofacial asymmetries in nonhuman primates by human judgments.