In order to evaluate the impact of a block in cancer medicine upon student interactions with cancer patients, a group of 32 students have been videotaped during interviews with simulated and genuine cancer patients before and after the block. The videotapes were scored and analyzed for frequencies and patterns of interactions by the method of Reciprocal Category Analysis. This disclosed significant decreases in purely factual questions by the students, with corresponding increases in discussion of emotional aspects of illness, following the course. The temporal pattern of interaction also changed with less of a tendency to delay or avoid addressing emotional aspects of disease noted. In addition to demonstrating changes in student interactions with cancer patients following a didactic course, this study exemplifies an approach offering more direct assessment of physician-patient interactions than has been practical using questionnaires.