Access to data about the clinical problems, patients, and processes that characterise family practice is essential for the development of this specialty. Practice-based research networks (PBRNs) play an increasing role in obtaining these data. We compared 3 PBRNs: one in Wisconsin in the United States, one in Wessex in the United Kingdom, and one in Nijmegen in the Netherlands. We organized our data into 4 key areas for review: the mission of the network, its contribution to the evidence base of family medicine, the management of the network, and the financing of the network infrastructure. Extending the evidence base of family practice is the overriding objective of these networks, and their main focus is on common morbidities. They provide access to unselected patient populations, but there are differences in their size. There are aspects of PBRNs that are common in countries with different health care systems, despite the fact that local circumstances--the research mission or the characteristics of the health care system under which they operate--determine their form and structure. Networks develop over time and their focus and activities may evolve. Financial support for these networks continues to be a problem.