STEINBACH and Hardison1 are correct when they state that the simple benefit-cost/utility (BCU) ratio must be refined for many practical decision situations, including the ones they describe. Because it did not affect the analysis of parasite screening among Indochinese refugees, the issue was not discussed in detail. But it is important, and bears directly on the General Health Policy Model's scope of applicability. This reply draws from the nonclinical literature where program or treatment interdependencies and nonlinear constraints have been studied in greater detail.2-5 A two-dimensional health policy space (Figure), where costs are subtracted from benefits (to establish consistency in signs), provides a useful analytic framework for considering resource allocation problems. Net dollar returns per person, where returns equals benefits minus costs, are plotted on the x axis, and well-years gained or lost per person (Y) are plotted on the y axis. Any alternative action (whether treatment, program,