The purpose of the research is to examine the effect of different response formats on student performance on introductory accounting exam questions. The study analyzes 1104 accounting students' responses to quantitative questions presented in two formats: multiple-choice and fill-in. Findings indicate that response format impacts student performance on most of the questions, in favor of the multiple-choice. A second goal is to discover how effectively wrong-answer options in multiple-choice questions capture students' actual incorrect thinking. Analyses of the fill-in questions showed students wrote in a wide array of incorrect answers, many of which would not have been offered as multiple-choice distracters. Answers filled in were often based on misguided logic that test question authors might never have anticipated. Students' incorrect reasoning on quantitative fill-in questions can inform instructors on how to proactively address student misconceptions and write more effective multiple-choice items.