In 2011, the digital library HathiTrust and partnering institutions approved a resolution to increase both access and coverage to U.S. federal government documents published since 1789. Experts estimate there are up to 2.2 million volumes—roughly equivalent to 500 million pages—of U.S. federal government documents. During the past two years, HathiTrust has worked to develop a comprehensive U.S. federal government documents registry. They have accumulated metadata records from a wide range of institutions. Yet the collected records are inconsistent and include an unknown number of duplicates. HathiTrust managers have identified duplicate records using a computer algorithm, but it is not perfect. We collaborated with HathiTrust to provide a method for analyzing the remaining unverifiable records. Our approach incorporates elements of crowdsourcing and gamification. The purpose of our project is twofold: to enhance the quality of the registry and engage a wide range of users.