Growth patterns of medical interviewing skills during a 6–year undergraduate cum'culum are assessed by studying 563 medical students taken from five year-groups, interviewing simulated patients. In a cross-sectional, quasi-experimental design their skills are rated by means of the Maastricht History-taking and Advice Checklist (MAAS), an observation instrument which measures five categories of interviewing skills pertaining to initial medical consultations. The findings suggest that the skills for‘history-taking’,‘presenting solutions’and‘structuring of the interview’are effectively learned. These learning effects result fiom a continuous small group teaching program with expert and peer review of videotaped encounters with simulated patients. The teaching effects of this program seem less for the skills pertinent to the phase of‘exploring the reasons for encounter’and to the‘basic interviewing skills’because the students' growing medical knowledge and the increasing ability to solve medical problems exert a counteracting influence on the acquisition of these easily deteriorating skills. The results might be helpjkl to cum'culum planners in order to make their programs for medical interviewing skills more effective.