Studies show that surgical face masks can have both positive and negative effects on attractiveness. Race has been implicated as a moderator of the size of this mask effect. Here, the moderating effects of expression, race and gender are explored. The mask effect was more positive for males than for females, for neutral faces than for smiling faces, and there were differences between the races. Further, the effect of unmasked attractiveness was partialled out for each image, which removed the race effects, but the gender and expression effects remained. It is suggested that racial differences previously observed in the mask effects are a consequence of differences in attractiveness of the faces sampled from those races. Re-analysis of previous research that showed race effects also demonstrates how they are better explained as attractiveness effects rather than race effects. This explanation can provide order to the different findings observed across the literature.