The study of sex differences is a rich, productive area of neuroscience, yielding findings that inform our understanding of basic biology and hold promise for clinical applications. There is a tremendous, problematic mismatch, however, between the actual implications of this research and what has generally been communicated to the public. The message communicated by the media, popular press, and in some cases researchers is often inaccurate with respect to what can and cannot be concluded from the data. This misrepresentation of findings has led to a crisis in public education and threatens to do the same in public health. Here, I suggest a number of ways that neuroscientists might address this growing problem. First, we should acknowledge that the term 'sex difference' is usually interpreted by the media and the public as evidence for dichotomous categories that do not actually exist. Because data rarely sort so cleanly into sex-specific categories, clearer presentation of the nature and size of sex differences is warranted. The term 'sex effect' may be preferable to 'sex difference' when the effect is not large. Second, factors that covary with sex, particularly experience, should be considered as causes of sex differences before the idea of "hardwiring" is invoked. Finally, we should be more vigilant about how our own findings are conveyed to policymakers and the public and speak out when they are misrepresented.