Establishing a link between mandibular morphology and diet in extant primates has long been a goal in biological anthropology because it should provide important insight into the diets of extinct primates, including fossil hominins. To date, efforts to explore this question have produced mixed results, largely perhaps due to a reliance on the use of 2D morphological data. Here, we report a study where we investigated whether 3D shape data would provide a clearer picture. We used geometric morphometrics to analyse 3D mandibular shape variation in a sample of > 200 primate specimens, representing individuals from 27 species and five families. Two sets of analyses investigated i) whether there was a relationship between mandibular shape and four standard dietary categories and ii) whether there was a relationship between mandibular shape and a well-known index of diet quality. We found an association between mandibular shape and the dietary categories when we employed raw Procrustes coordinates and allometry-free residuals, but the relationship was weak to non-existent when the effects of phylogeny were taken into account. We found no relationship between shape and the diet quality index, no matter whether the data were raw, corrected for the effects of allometry, corrected for the effects of phylogeny, or corrected for the effects of both allometry and phylogeny. Taken together, the results of the two sets of analyses suggest that there is a weak relationship between 3D mandibular shape and diet in extant primates. Allometry and phylogeny appear to be more important influences on the 3D shape of extant primate mandibles than is diet. We conclude from this that 3D analysis of mandibular shape is unlikely to further illuminate the diets of extinct primates, and research efforts should, therefore, be directed elsewhere.