In this study we observed college students' behavior on two self-tailored, multiple-choice exams. Self-tailoring was defined as an option to omit up to five items from being scored on an exam. Participants, 80 undergraduate college students enrolled in two sections of an educational psychology course, statistically significantly improved their scores on both exams by using the self-tailoring option. They omitted a statistically significantly higher number of correctly answered items on the second exam than on the first exam. For both exams, a statistically significant negative correlation was present between how frequently items were omitted and item difficulty values; students were more likely to omit more difficult items from scoring than less difficult items. (Contains 2 tables.)