Keyword and activities list approaches to measuring women's work are compared. The two approaches were applied to the same population of women in Egypt in two consecutive surveys. The widely used keyword approach underestimates women's work rates, disproportionately excluding poor and poorly educated women, particularly those working in nonformal jobs. The activities list approach captures these missed economic activities and also the multiple jobs women hold simultaneously. Survey measurement of women's work must be improved to fully account for women's contributions to economic life and to better understand the relationship of work to such other important variables and processes as reproductive change, child welfare, and economic development.