The geographical distribution of active physicians who graduated from medical schools before 1985 were studied. Those who had emigrated, hold non-medical jobs, are in the military service, or work as public health physicians and resident staff were excluded from the study. A total of 27,728 physicians were analyzed. Our studies have shown a relationship between the location of the medical schools from the which the physicians have graduated and the geographical regions in which they practice. A statistically significant number of physicians are working near the medical colleges from which they have graduated. That is, those who had graduated from medical schools located in the southern area of the country are presently working in the same region. This relationship was shown to be especially significant for older physicians and female doctors, who work around the area of the medical colleges from which they graduated.
In order to compare the geographical distribution of physician by level of medical care and specialty, a log linear model was applied to the annual registration data of the Korean Medical Association as of the end of December, 1991 which was supplemented from related institutions and adjusted with relevant sources. Those physicians in primary and secondary care institutions were not statistically significantly unevenly distributed by province-level catchment area. There were some differences in physician distribution among big cities, medium and small-sized cities, and counties; however, those physicians for primary care level were equitably distributed between cities and counties. Specialties for secondary care physicians were less evenly distributed in county areas than in city areas, and generalists are distributed more evenly in cities and counties than in big cities. There is a certain limitation due to underregistration in the annual physician registration to the Korean Medical Association; however, the geographical distribution of physicians has been improved quantitatively. It is strongly suggested that specialties and the level of medical care should be considered for further physician manpower studies.