Studies of cooperatively breeding birds rarely benefit from the extensive research on adaptive foraging behaviour, despite the potential for concepts such as state-dependent foraging to explain many aspects of behaviour in social groups. For example, sex differences in preferred foraging techniques used by green woodhoopoes, Phoeniculus purpureus, have previously been explained by sexual dimorphism in bill length and the benefits afforded by foraging specialization and niche differentiation within cooperative groups. Contrary to this argument, there were no sex differences in mean foraging success and/or prey size captured when males and females used the same foraging techniques. Subordinates of both sexes did experience lower and more varied foraging success compared with dominants, but probably only as a consequence of competition or inexperience. However, dominant males experienced greater variance in individual foraging success compared with dominant females, and dominant males also experienced greater variances in prey size when using their preferred foraging techniques. Dominant males therefore appeared to specialize in foraging techniques that provided more variable rewards, whilst dominant females consistently chose to minimize variation in reward. Dominant females also experienced less variance in foraging returns when using the same techniques as males, suggesting a possible link with sexual dimorphism in bill length. Partitioning of foraging niches in dominant green woodhoopoes therefore appears to be better explained by sex differences in variance (risk) sensitivity to foraging rewards. We suggest that this kind of detailed analysis of state-dependent foraging has the potential to explain many of the crucial age and sex differences in behaviour within cooperative groups.