The extent and benefit of answer switching when completing multiple-choice exams was investigatedin an undergraduate course (n = 1,152) and a postgraduate course (n = 1,624). Answerswitching was identified in 1.7% and 2.4% of cases, respectively. In both samples, more than halfthe participants changed at least one answer and, of these, approximately 50% increased their testscore and 25% decreased their test score. Significant gender differences were not found, but maleswere less likely to switch. Multivariate analysis indicated no significant differences in answer switching behaviour between Australian and international students. Univariate tests, however, suggested that international students made more right to wrong (p <.05) and wrong to wrong (p <.02) switches. The results also suggested that better students were more likely to switch from a wrong to a right answer, and to make significantly fewer right to wrong (p < .001) switches.