Intergroup aggression is relatively common in many primate species and its functions have long been debated and incorporated as important parts of primate socioecological models. Recent studies have improved upon previous research by empirically testing the potential mate, food/resource and infant defence functions of intergroup aggression. I raise three main issues regarding the predictions and methods that have been used and make suggestions that should improve future studies. First, I address what constitutes evidence of defence and suggest that encounter rates and contexts should generally not be used to help test mate, resource and infant defence hypotheses. Secondly, I address the problem of disentangling mate, resource and infant defence hypotheses and suggest that because they are not mutually exclusive and potentially confound one another, tests of one hypothesis should adequately control for other possible functions, preferably using multiple logistic regression. Lastly, I discuss the importance of examining intergroup aggression using a cost-benefit approach and suggest that measures should be taken to control for species-specific measures of fighting ability, and/or other potential costs or influential factors that might influence participation in intergroup aggression, including groups' identities. For the latter two issues, I provide simulations that illustrate the methodologies I suggest and the potential effects of employing them, relative to previously employed tests.