Despite considerable evidence documenting a strong and persistent relationship between socioeconomic position and mortality, recent research suggests that this association may be weaker among women. In our examination of gender differences in the socioeconomic gradient in mortality, we argue that this inconsistency arises from the failure to consider the ways in which gender is a fundamental constituent of socioeconomic position. The data used are from the Panel Study of Income Dynamics. Respondents, including all household heads and their partners, aged 29 years and older in 1972 (N = 5,665; 56% female), were followed until 1991, death, or attrition. Discrete time event history analysis was used to examine the predictors of death between 1972 and 1991. Of the key socioeconomic predictors, years of education was measured at baseline, while earned income was a time-varying covariate. We find no gender differences in the effect of respondents' own socioeconomic positions on their mortality risk. However, increasing spousal income raises men's odds of dying, while the opposite is true for women. Our results raise questions about the prevailing view that the socioeconomic gradient in mortality is weaker among women. Moreover, gender differences in the effects of spousal earnings on mortality risk suggest that their labor market rewards have fundamentally different meanings for women and men.