The goal of the current research was to gain insight on how people visualize, or “picture in their heads”, the faces of immigrants, and what role anti-immigrant biases may contribute to these facial representations. Participants (image generators) completed a reverse-correlation image-classification task in which they generated a facial image of an immigrant or a natural-born American citizen. Separate groups of participants (image raters) then evaluated these facial images on various traits (e.g., competence, trustworthiness, dominance), or classified them by perceived race/ethnicity. Results revealed that the immigrant facial representation were rated more negatively and were more likely to be classified as non-White, relative to the citizen images. These differences were more pronounced in the visualizations created by image generators who had less positive immigrant attitudes (as assessed by a Single-Category Implicit Association Test). Overall, these findings suggest that anti-immigrant biases may shape the way in which immigrants are visualized.