Extraordinarily high rates of recombination have been observed in some eusocial species. The most popular explanation is that increased recombination increases genetic variation among workers, which in turn increases colony performance, for example by increasing parasite resistance. However, support for the generality of higher recombination rates among eusocial organisms remains weak, due to low sample size and a lack of phylogenetic independence of observations. Recombination rate, although difficult to measure directly, is correlated with chromosome number. As predicted, several authors have noted that chromosome numbers are higher among the eusocial species of Hymenoptera (ants, bees and wasps). Here, we present a formal comparative analysis of karyotype data from 1567 species of Hymenoptera. Contrary to earlier studies, we find no evidence for an absolute difference between chromosome number in eusocial and solitary species of Hymenoptera. However, we find support for an increased rate of chromosome number change in eusocial taxa. We show that among eusocial taxa colony size is able to explain some of the variation in chromosome number: intermediate-sized colonies have more chromosomes than those that are either very small or very large. However, we were unable to detect effects of a number of other colony characteristics predicted to affect recombination rate – including colony relatedness and caste number. Taken together, our results support the view that a eusocial lifestyle has led to variable selection pressure for increased recombination rates, but that identifying the factors contributing to this variable selection will require further theoretical and empirical effort.