Sex-specific cognitive abilities are well documented. These can occur when sexes engage in different ecological contexts. Less known is whether different ecological contexts can also drive sex-specific participation rates in behavioral tests. Here, we explore this question in bumble bees, a group of eusocial insects where worker females and males exhibit stark socioecological differences. Among myriad colony maintenance tasks, workers forage for themselves and developing brood, while males forage only for themselves while mate-searching. Following upon previous studies suggesting no sex differences in bumble bee learning, we test the hypothesis that despite having equivalent associative learning abilities, males participate in cognitive assessments offering nutritional rewards at lower rates. Testing >500 bees from nine colonies in a differential conditioning protocol, we find support for our hypothesis. An equivalent proportion of workers and males successfully completed our cognitive assessment, while a significantly lower proportion of males participated in the entire protocol. Unequal participation is a perennial issue in the behavioral sciences, limiting sample size and potentially biasing results. Our results suggest that to understand the true range of variation in cognition, sex-differences in participation must be accounted for.