Background: The Access and Delivery Services and Collection Development departments at the Cushing/Whitney Medical Library have a cooperative relationship in terms of incorporating statistical ILL data into collection decisions. Budget pressures resulting in journal cancellations have made the need for data-driven decision making more acute. A question arose as to how accurately ILL statistics for journals reflect scholars’ needs in relation to library collection building. Objective: The researchers hope to determine if borrowing statistics after the cancellation of a subscribed journal reflected the same level of use compared to prior to cancellation. Methods: The researchers gathered ILL borrowing statistics for a control list of titles cancelled in fiscal year 2017/2018 and analyzed them against COUNTER JR1 statistics for the same period of time to determine if interest in a journal remained consistent post-cancellation. Results: This study’s preliminary results indicate that ILL borrowing statistics are significantly lower than anticipated based on documented usage. These results indicate that potential users are not taking additional steps to obtain articles through ILL that the library no longer subscribes to. Outcome: Statistical data with no context cannot reliably capture user behavior and librarians should consider the incomplete nature of these metrics when anticipating user need.