Usability testing is a key step in the successful design of new technologies and tools, ensuring that heterogeneous populations will be able to interact easily with innovative applications. While usability testing methods of productivity tools (e.g., text editors, spreadsheets, or management tools) are varied, widely available, and valuable, analyzing the usability of games, especially educational “serious” games, presents unique usability challenges. Because games are fundamentally different than general productivity tools, “traditional” usability instruments valid for productivity applications may fall short when used for serious games. In this work we present a methodology especially designed to facilitate usability testing for serious games, taking into account the specific needs of such applications and resulting in a systematically produced list of suggested improvements from large amounts of recorded gameplay data. This methodology was applied to a case study for a medical educational game, MasterMed, intended to improve patients’ medication knowledge. We present the results from this methodology applied to MasterMed and a summary of the central lessons learned that are likely useful for researchers who aim to tune and improve their own serious games before releasing them for the general public.