External validity is the problem of generalizing results from laboratory to non‐laboratory conditions. In this paper we review various ways in which the problem can be tackled, depending on the kind of experiment one is doing. Using a concrete example, we highlight in particular the distinction between external validity and robustness, and point out that many experiments are not aimed at a well‐specified real‐world target but rather contribute to a ‘library of robust phenomena’, a body of experimental knowledge to be applied case by case.