Undergraduates studied the definitions of 16 psychology terms, expecting either a multiple-choice (n = 132) or short-answer (n = 122) test. All students then received the same multiple-choice test, requiring them to recognize the definitions as well as novel examples of the terms. Compared to students expecting a multiple-choice test, those expecting a short-answer test performed similarly on example questions but significantly better on definition questions. Students in these 2 test-expectation conditions also differed in several subjective ratings of their study and test taking. The results suggest that students do not typically study in an optimal way for multiple-choice tests. (Contains 1 table and 2 notes.)