The military services of the United States maintain an extensive health care delivery system in order to ensure the appropriate level and availability of care to the active duty forces. If only the active duty personnel were to use these facilities they would operate at only a fraction of that possible given the necessity to staff for the military contingency plans. Thus, given the expansion of the health care fringe benefit package of the active duty and retired personnel, the non-active duty population for whose care the military become responsible in one form or another have been allowed, and sometimes urged to utilize at least a portion of this excess system capacity. The end of the draft and the resulting need to compete in the marketplace for medical personnel, as well as the general inflation in the health care sector, has spotlighted the increasing cost of caring for these dependent groups. The question has arisen of whether it might not be cheaper to shift some of this demand for health care to the civilian sector. In this paper we examine analytically the appropriate considerations and elements to be compared in this research point out the crucial empirical work necessary to estimate such a model, discover some of the ways in which the analytical construct can provid3e bounds and directions to the hypotheses to be tested, and finally conjecture some preliminary policy recommendations. (Author)