The Rab GTPases are the largest family of small GTPases, with 66 members in humans, and play essential roles in membrane trafficking by recruiting effector proteins to the correct membrane at the correct time. Indeed, many proteins involved in membrane trafficking form direct interactions with Rab GTPases; thus, identifying proteins which interact with Rabs can provide useful insights to the mechanisms underlying protein trafficking processes. Recently, AlphaFold has emerged as a useful tool for not only modelling the structures of known protein complexes, but also for the identification of novel interactions between proteins. In this work, we have applied AlphaFold2-Multimer modelling to perform a large-scale, in silico screen for Rab interacting partners. Using hits from published proteomics datasets, we tested 676 potential Rab interactors, of which 121 (17.8%) could be modelled in complex with a Rab with high confidence. This includes both known Rab interactors and those which are unreported in the literature. Notably, our modelling has revealed that several Syntaxin proteins (which act as SNAREs in membrane fusion) may interact with Rab proteins. Two of these interactions predicted by AlphaFold—the Syntaxin1A-Rab6B interaction and the Syntaxin7-Rab14 interaction—have been validated in vitro using isothermal titration calorimetry. Furthermore, we have found that the BEACH domain, a highly conserved domain of unknown function, interacts with Rab proteins. Surprisingly, both our AlphaFold models and in vitro assays suggest that the BEACH domain functions as a GTPase-activating protein (GAP); that is, it functions to inactivate Rab proteins by stimulating GTP hydrolysis. Overall, these findings demonstrate that pairing AlphaFold with proteomics datasets is a powerful method for the discovery of novel protein-protein interactions.