Data are presented from the Medical Care Group of Washington University and from several other organized medical care settings responsible for prepaid enrollees. As few as 8 per cent of visit may be made to specialists under systems which are organized to emphazide primary care. Further, the proportions of primary care and specialty visits vary inversely weith the "strictness" of this organization. Those groups which employ family physicians have a higher proportion of primary care physicians than do those which employ general intermists and pediatricans, or than exist in the medical care system at large. The data suggest strongly that the organization of medical care, as well as the training of personnel employed has a major influence upon the types of health manpower needed. Future research in this area is urgent, but even the currently available data have important implications for health manpower policy.