Field evidence suggests that people belonging to the same group often behave similarly, i.e., behaviour exhibits social interaction effects. We conduct an experiment that avoids the identification problem present in the field. Our novel design feature is that each subject simultaneously is a member of two randomly assigned and identical groups where only members (‘neighbours’) are different. In both groups subjects contribute to a public good. We speak of social interactions if the same subject at the same time makes group-specific contributions that depend on their respective neighbours’ contribution. We find that a majority of subjects exhibits social interaction effects.