We use data from the Health and Retirement Study to examine the role of work-life balance (WLB) as a nonmonetary determinant of retirement transitions, conditional on job attributes such as hours of work, compensation and benefits. We show that low levels of WLB are significantly associated with subsequent reductions in labor supply for workers aged 51 to 79, and document heterogeneity by gender and employment status. Moreover, WLB mediates labor supply responses to spousal health shocks. Workers who report higher levels of work-to-life interference are significantly more likely to reduce their labor supply in the next two periods following a spouse’s health shock, and this effect is once more heterogeneous. The moderating effect of WLB is stronger for women than men. Among female workers, it is stronger for those employed part-time at baseline.