In this study conducted with 80 first-year students in a graduate medical course at the Australian National University, Canberra, students' critical thinking skills were assessed using the Watson-Glaser Critical Thinking Appraisal (Forms A and B) in a test-retest design. Results suggested that overall subjects retained consistent patterns of ability in critical thinking over the year. Subjects aged 30 years and over were more likely to perform better on the tests than subjects aged under 30. However, there was a significantly lower group mean on Form B than on Form A. Subjects with an above-mean score on Form A, or subjects aged 30 years and over, were more likely to be associated with a reduced score on Form B than were subjects with a below-mean score on Form A, or aged less than 30 years. In motivational terms, it is suggested that subjects who had more highly developed critical thinking skills, and older subjects, conducted a more realistic effort/reward analysis before the administration of Form B, and were less willing to put effort into a test that was not an assessable component of their medical school curriculum, than were other students. For the design of assessment items in higher education, this would suggest that assessment weightings need to reflect accurately the size of the reward, or possible outcome, of completing those assessments if students are to work to their ability. (Contains 7 tables and 3 figures.)