A physical inventory of the open stacks collection at Indiana University was conducted to determine the rate of error in the corresponding bibliographic records. The inventory was started to address some errors that were found when materials were pulled for offsite storage, but took on an increased importance as participation in shared print programs increased. This article describes the methodology used to conduct the inventory, as well as the rate of error found in the records for a large, open stack collection. This is an Accepted Manuscript of an article published by Taylor & Francis in Collection Management on September 9, 2020 available online: https://www.tandfonline.com/doi/10.1080/01462679.2020.1818343 Inventory dataset available at: http://hdl.handle.net/20.500.12419/592