A method is presented to test the relationship between a phylogenetic tree derived from brain morphology, and different hypotheses describing the evolution of a behavioral trait. This is a question of interest for evolutionary psychologists and behavioral biologists. The paper first discusses how hypotheses for behavioral evolution should be coded for such a comparison, then a triple-per-mutation test, originally proposed to compare independently obtained evolutionary trees, is used for the statistical assessment of each hypothesis. Non-parametric correlation coefficients computed between brain components and appropriately coded behavioral states can then be used to suggest what brain components are responsible for the development of the various states of the behavioral trait of interest. The procedure is illustrated with three different applications relating brain evolution to habitat selection in marsupials, locomotory specialization in primates, and trophic adaptation in bats.