Ambivalent sexism is a two-dimensional framework that assesses sexist and misogynous attitudes. The current corpus of research on such attitudes suggest that they are predicted by numerous variables, including religious beliefs, ideological variables, and men's facial hair. Most studies, however, have treated such predictors as if they are independent – inferring that zero-order correlations between sexism and its predictors are not confounded by omitted third variables. In the current work, we address ambivalent sexism using a large array of known correlates of sexist attitudes in two large and demographically diverse samples. We show that low empathic concern is the primary driver of hostile-, but not benevolent sexism (Study 1); that social dominance orientation, right-wing authoritarianism, religiosity, and low Openness and Agreeableness differentially predict ambivalent sexism (Study 2); along with male gender and low education level (Study 1 and 2). Contradicting an earlier finding, men's facial hair was not correlated with hostile sexism in either studies and a short full beard predicted lower scores on benevolent sexism in Study 2. Thus, we replicated the main findings from most previous research except for men's facial hair, and we also show the paths through which predictors of sexist attitudes exert their effects.