Multicenter studies and research in primary care practice-based research networks are increasingly important in identifying the most cost-effective interventions for a variety of health problems. However, such research has necessitated multiple institutional review boards (IRBs) to review proposals, using local interpretation of the federal requirements governing the conduct of research involving human subjects. Finch and colleagues examined the process and outcomes of local IRB review by 75 IRBs of 2 research network studies that had been reviewed and approved by university IRBs. Practices requiring additional local IRB approval agreed to participate less than those that did not. Of the 88 practices requiring local IRB approval, 55 received approval, with nearly 50% needing active help by the research network, many requiring consent changes. Practices requiring local IRB approval were less likely to complete data collection but more likely to enroll minority patients. When considering future reforms, the national human subject protections system should consider the potential redundancy and effect on generalizability, particularly regarding enrollment of poor urban children related to local IRB review.