The spatial distribution of grain legumes can be explained by combining the analysis of traditional location factors with the concept of spatial dependence. We present an in-depth literature review and derive hypotheses about the conditions that make a certain location attractive for grain legume cultivation. These hypotheses are then tested via three different approaches: a fractional logit generalized linear regression model, a zero-one inflated beta regression model and spatial econometric models. We use secondary data for the EU-28 at the NUTS2 level. Location factors that contribute to the incidence of grain legume cultivation are the frequency of organic farming and irrigable area, a farther distance to the next main port, a prominent role of legumes in regional diets, and policy measures such as support coupled to protein crops and restrictions on the use of nitrogen fertilizers. Our results suggest that agglomeration effects also matter in the cultivation of dry pulses. Corresponding external economies of scale and positive spill-over effects could be further exploited by specifically supporting investments in regional supply chains, local extension services and training.