Data from the literature on the baculum and our own studies are used to evaluate two conflicting hypotheses for bacular variation. The first hypothesis states that bacular differences arise incidentally as pleiotropic by-products of phyletic divergence. The second hypothesis attributes precise reproductive function to the baculum (and phallus), which in turn allows the baculum to function as a reproductive isolating mechanism. Several patterns of bacular variation are used to test these hypotheses: bacular length versus body size (both among and within taxa), baculum length versus vaginal length, and bacular variability versus baubellar (os clitoris) variability. The results of these analyses lead us to propose that, by determining gross phallic morphology, different bacular morphologies probably elicit different behavioral and neuroendocrine responses from the female during copulation. Changes in bacular morphology that produce differential responsiveness on the part of the female may thus provide a basis for species-specific reproduction. Additional analyses of phyletic and geographical patterns of bacular variation corroborate this proposal, and the systematic and reproductive implications of these results are briefly considered.