This article utilizes benefit-cost analysis to evaluate a hypothetical rural hospital closure. The estimated benefits were found to be extremely sensitive to assumptions concerning the nature of hospital marginal cost curves and less sensitive to the proportion of the hospital bill assumed to be paid directly by consumers. The net benefits of a hospital closure were largest in the case where hospital alternatives were small, under-utilized facilities. The analysis suggests that when evaluating medical delivery system change, benefit-cost techniques can be useful to policymakers in structuring the problem and identifying key assumptions. However, theoretical considerations in combination with data and resource requirements will limit its usefulness as an applied health-planning tool.