As the gap between retirement resources and needs grows, many researchers have prescribed the antidote of working longer. But this prescription may disadvantage lower socioeconomic status (SES) households because they have shorter lives than higher-SES households, and working longer may increase existing disparities in retirement durations. This paper uses data from the National Longitudinal Mortality Study to quantity these disparities by SES since 1979, using education as a proxy for SES. The paper finds that age-65 life expectancies have increased for all levels of educational attainment but that the gains have been much greater for those in the top quartile. The paper uses these calculations to determine hypothetical retirement ages that hold constant for each SES group the 1979 ratios of time spent in retirement to time spent working. The findings suggest that all educational groups can work longer today than in the past, while spending a similar fraction of their lives in retirement; those in the top quartile of educational attainment can work a full one to two years longer than those in the bottom quartile and still maintain their 1979 ratios.