Many researchers have questioned the validity of the underlying assumptions, the psychometric properties, and conceptualizations for both the older and new measures that generate two independent scores for masculinity and femininity. The assumptions of bipolarity and unidimensionality and the construct validity of existing masculinity/femininity inventories have been identified as the main areas of criticism. An early-adolescent sample of 418 junior high students and a post-adolescent population of 442 college students were given the Femininity Scale from the California Psychological Inventory and the Bem Sex Role Inventory to examine the psychometric properties and underlying assumptions of these two different measures of masculinity and femininity, as well as the utility and interpretation of each measure. Comparisons were made between the two tests and between age groups. Factor analyses suggested that both tests were multidimensional. The bipolarity of masculinity-femininity was not supported. Sex and age differences were found in the factor pattern and correlations with factors. Conceptual problems inherent in the constructs of masculinity and femininity because of present empirical definitions suggest a need for changes in terminology and assessment approaches. (Author)