Grounded in the theory of ambivalent sexism, this study tested the speculation that women's benevolent sexist attitudes may be, in part, a self-protective response to environments they perceive as hostile to women. Data that have indirectly supported this conjecture thus far have been correlational. The current study involved a more powerful, experimental test of the hypothesis. Women (N=105) were randomly assigned to one of three conditions, which differed only in what participants were told about research findings on men's attitudes toward women (negative or positive attitudes, or no information). As predicted, benevolent sexist attitudes--but not hostile sexist attitudes--were strongest for women told that men hold negative attitudes toward women. This effect is consistent with a benevolent sexism-as-protest explanation and was statistically significant even while controlling for attitudes toward feminism. The differential effect of beliefs about men's attitudes on these two types of sexism lends further support to the idea that, although hostile and benevolent sexism are related, they may serve different functions.