AbstractEmpirical studies of bilateral foreign direct investment (FDI) activity show substantial differences in specifications with little agreement on the set of included covariates. We use Bayesian statistical techniques that allow one to select from a large set of candidates those variables most likely to be determinants of FDI activity. The variables with consistently high inclusion probabilities include traditional gravity variables, cultural distance factors, relative labour endowments and trade agreements. There is little support for multilateral trade openness, most host‐country business costs, host‐country infrastructure and host‐country institutions. Our results suggest that many covariates found significant by previous studies are not robust.