Despite well-intentioned gender diversity initiatives aimed at addressing gender imbalances by ensuring minimal female representation in predominantly male groups, such tokenism often exacerbates discrimination and social isolation for these women, potentially leading to absenteeism. Research suggests that the benefits of diversity are realized only when the ratio of women to men reaches a critical threshold that allows for genuine integration and participation. However, this threshold remains uncertain. We integrate tokenism theory with social identity and status characteristics theories to investigate the effects of gender ratios within organizational teams on individual absenteeism. Specifically, we theorize a U-shaped relationship between gender dissimilarity and absenteeism for women, but not for men. Study 1, with a one-year cross-lagged design, encompassing 10,332 blue-collar workers in 1064 teams, supports the U-shaped relationship for women, while the relationship for men was non-significant. In Study 2, we use an experimental design with a sample of 370 female blue-collar workers to explore two potential mechanisms that may together explain the U-shaped gender dissimilarity effect for women. We test whether the gender composition of the work group affects both women’s likelihood of reporting unpleasant experiences and the group’s norms regarding absence. We draw theoretical and practical implications from these findings.