Voluntary meetings with costly participation are ubiquitous as an institution for resource allocation. Despite their importance in economic life, such meetings are little studied. This paper is an empirical analysis of participation at public meetings. We investigate basic and previously unaddressed questions about meetings with costly participation. Who goes? Does attendance vary with observable characteristics? Do meeting attendees represent the interested population? We find that the opinions of participants do not represent the opinions of the entire regulated population and that small changes to meeting protocols have the potential to manipulate the sample of participants. We also find surprisingly strong evidence that the opinions of participants are extreme relative to the whole population. These results lay a foundation for the problem of tailoring meeting protocols to achieve particular welfare objectives.