The subjects (N = 175), freshmen and fifth year students in psychology (n = 59) and medicine (n = 116), described their modes of studying by answering open-ended questions and Likert-type questions, presented to them within a task booklet. They were first asked to write down their own subjective definitions of learning. Scales to classify the answers were adopted from Lonka et al. (1990). Second, subjects were given a comprehension monitoring probe adopted from Ryan (1984). Responses were analysed to determine the specific comprehension criteria the student employed. Third, students rated a set of 71 statements concerning learning styles, regulation of learning, and conceptions of learning (Entwistle & Ramsden, 1983;Vermunt & van Rijswijk, 1988). Embedded in the last set of statements were seven items that were used to classify students as dualists or relativists (Perry, 1968;Ryan, 1984). It was found that constructivist conceptions of learning were the most typical of (advanced) psychology students, whereas learning was more often seen as intake of knowledge by the medical students. Highest dualism scores were obtained by the first year students, especially medical students. In general, dualists were more likely to report knowledge comprehension criteria to test their understanding than were relativists, and dualists' conceptions of learning were also more passive. Four principal components were identified that reflected qualitatively different approaches to learning and knowledge: externally regulated and reproduction-directed learning (PC 1), self-regulated, meaning-directed, and goal-oriented learning (PC2), constructivist epistemology (PC3), and active professional orientation (PC4). Medical students scored higher on variables related to PC1 and PC4, whereas psychology students scored higher on scales associated with PC2 and PC3.