For educational achievement tests that employ multiple-choice (MC) items and aim to reliably classify students into performance categories, it is critical to design MC items that are capable of discriminating student performance according to the stated achievement levels. This is accomplished, in part, by clearly understanding how item design features align with each of the achievement level descriptions. Difficulty drivers are defined as specific item features that are related to the difficulty of an item. The purpose of this research study was to pilot a methodology for working with subject matter experts to identify what item features drive item difficulty, and to determine specific features that drive item difficulty on a summative large scale assessment used in high school social science and history classrooms. Results of the study indicated that although this methodology was useful for gathering such information, modifications could be made in the future to better elicit the desired information about difficulty drivers. [Slides presented at the Annual Meeting of the American Educational Research Association (AERA) and National Council on Measurement in Education (NCME) in New Orleans, LA in April 2011.]