There has been much interest in studying evolutionary games in structured populations, often modelled as graphs. However, most analytical results so far have only been obtained for two-player or additive games, while the study of more complex multiplayer games has been usually tackled by computer simulations. Here we investigate evolutionary multiplayer games in regular graphs updated with a Moran process. Using a combination of pair approximation and diffusion approximation, we obtain an analytical condition for cooperation to be favored by natural selection, given in terms of the payoffs of the game and a set of structure coefficients. We show that, for a large class of cooperative dilemmas, graph-structured populations are stronger promoters of cooperation than populations lacking spatial structure. Computer simulations validate our results, showing that the complexity arising from many-person social interactions and spatial structure can be often captured by analytical methods.