To investigate the idea that providing information about a job applicant's past performance can avert sex discrimination in preliminary employment decisions, an experiment was conducted in which both Applicant Sex and Type of Information were varied. As predicted, highly job-relevant information was found to produce less differential treatment of male and female applicants than did information of low job relevance or no information at all. Also as predicted, the type of information provided had more impact on reactions to female applicants than male applicants, with high job-relevance information producing the most favorable responses and, unexpectedly, low job-relevance information producing the least favorable responses to female applicants. Additional results suggested that these effects were mediated by the degree to which female job applicants were characterized by stereotypic attributes. The findings are interpreted as supportive of the idea that undermining the information value of sex stereotypes as a basis of inference about the attributes of a given woman can function to reduce sex discrimination in employment settings.