BACKGROUND The questions raised in this article is whether the social background of Norwegian medical students has changed over time, and whether changes in admission procedures have influenced recruitment. METHODS The analysis is based on the total population of Norwegian students who qualified for university admission between 1980 and 2003, including those who entered medical school. The proportion of medical students who qualified for admission to medicine through private-school tutoring is estimated for the whole period, so is the degree to which medical students read other university subjects prior to entering medical school in order to get in. The impact of social origin is studied by estimates of the proportions of university-qualified students from various social strata who went to medical school, as well as by logistic regression analysis. RESULTS The use of private-school tutoring has decreased in recent years, while the proportion who read other university subject prior to admission has increased. The impact of social background has not changed. Those who have physicians as parents are particularly likely to go to medical school; the relative difference between them and students with parents with other university degrees or a lower educational attainments tends to be stable over time. This also holds true when we compare students with the same grade level in school. CONCLUSIONS Three explanations are offered for the stable social recruitment, and especially for the high proportion of physicians' children recruited: physicians have ample financial means to support their children, their children develop an interest in medicine because of their parents' line of work, and they wish to avoid social degradation.