Poverty measurement is extremely important, but most contemporary measures of poverty are deeply flawed.  One major shortcoming of all existing poverty measures is a lack of gender sensitivity, which includes but is not limited to an inability to reveal the gendered distribution of deprivation, particularly for the worst off.  In this paper I provide a methodological framework for thinking about gender and poverty measurement, and suggest how a gender sensitive, pro-poor methodology can  inform both the procedures for developing new poverty measures and the substance of those measures.