The overarching purpose of this dissertation is to provide the literature with a revised conceptualization of gender beliefs that will better explain variance in public opinion. When trying to operationalize feminist attitudes in the past, the public opinion literature has relied on measures of abortion or one-dimensional index scores- basically collapsing attitudes into a false dichotomy of feminist/anti-feminist. This is problematic for many reasons. I argue that the feminist belief system should be treated as a multidimensional concept comprising at least three distinct dimensions: belief about women's opportunity; belief about assertive women; and belief about the changing family structure and role of the mother. The second half of the dissertation applies this new approach within the areas of abortion attitudes, candidate evaluations (both experimental and real world), and party affiliation. Overall, the findings support the thesis that there are at least three distinct gender beliefs with varying degrees of explanatory power.