BACKGROUND: To facilitate students' transition from basic, science-oriented, problem-based learning (PBL) to clinical reasoning-oriented PBL, the University of Geneva School of Medicine introduced a 12-week unit of Introduction to Clinical Reasoning (ICR) at the beginning of its fourth or clerkship year. PURPOSE: The aims of the present study were to determine, after 12 weeks in the ICR unit, to what extent students had: (1) identified the learning content set by the faculty while adapting to the hypothetico-deductive reasoning approach; (2) familiarised themselves with the clinical reasoning-oriented learning process, and (3) transferred and further developed this process during the clinical years. METHOD: Students' derived objectives from the problems were compared to the objectives preset by the faculty to determine acquisition of intended learning content. To assess their adaptation to the clinical reasoning-oriented PBL approach, students (n = 124) were asked to list and freely comment on aspects of the unit they felt most at ease with or had difficulty with, and to complete a questionnaire on the clinical reasoning process (CRP). The same questionnaire was administered 6 and 12 months later to assess the evolution of the students' self-perception during clerkships. RESULTS: On average, student objectives matched 62% of faculty objectives. Half of the missed (38%) objectives were in basic sciences. Students generated 16% additional objectives, also predominantly in the basic sciences category (41%). Free comments indicated that the difficulties perceived by students were very similar to those previously reported in studies on reasoning and errors, such as difficulty in gathering, interpreting and weighting relevant data, synthesising information, and organising it hierarchically. These results were confirmed with the CRP questionnaire administered at the end of the unit. For most of the competencies assessed on the CRP questionnaire, a gradual improvement was seen to have occurred by 6 and 12 months after the unit. CONCLUSIONS: To ease students' transition from the preclinical to clinical years, a learning unit should give them the opportunity to train their clinical reasoning processes on standardised and prototypical problems, before encountering real patients with more ill-structured problems during clerkships. Such a transitional structure should particularly emphasise a developed repertoire of problem representations, recognition of key findings and a hierarchical classification of working hypotheses. It should foster the creation of links between the acquired basic clinical knowledge and the diagnostic, management and therapy steps of problem solving.