A methodology is developed to address two key related questions for the Engineering Test Reactor (ETR): 1. How do we make the best rational technical decision in selecting the optimum ETR design? 2. What is the uncertainty in this design point due to uncertainties in the system models and input data base we employ in formulating the design? The methodology provides a framework to reduce complex decision problems for the ETR to a quantitative form that compels us to consider all relevant technical factors and could form the basis of a consensus process for ETR design selection. The first key element in this process is the formulation of a utility function, a combination of ETR attributes (e.g., capital cost, wall loading, ignition margin, design credibility, etc.) into a single number that can be used as a global figure of merit to express the worth of a particular design and can be used as a value function for the intercomparison and optimization of ETR design concepts. In particular, we emphasize that intangible attributes, such as “reactor relevance” or “risk/confidence in filling mission goals” can be incorporated in a systematic manner. Next, a procedure is developed for determining the optimum design point from a field of available options, with concurrent acknowledgment of the uncertainty in the design point. We demonstrate that this involves the selection of the best set of “internal” systems variables (i.e., variables that we are free to specify in the design, such as major radius, plasma current, magnet current density, etc.) and an appropriate set of “external” systems variables (i.e., variables that are determined by external means and not under our control but that have associated uncertainties, such as the Troyon beta scaling coefficient, the unit cost of a plasma heating system, etc.), in order to maximize the value of the utility function and determine its uncertainty. We then expand the discussion of uncertainty through sensitivity analysis of the ETR design to emphasize that, because we have never yet experimentally observed a deuterium-tritium burning tokamak with high 14-MeV neutron fluxes and sustained by alpha-particle heating, assessment of the uncertainty in our design can be as important as the design point itself. We proceed to show that this uncertainty can be expressed as the quadrature combination of sensitivity profiles and individual uncertainties for all external systems variables employed in formulating the design and provide some illustrative examples for the current Tokamak Ignition/Burn Experimental Reactor ETR design due to uncertainties in the modeling data base. Finally, we demonstrate that our formalism is ideally suited for pinpointing critical external systems variables and models that must be determined to a greater accuracy to reduce the overall ETR design uncertainty to acceptable levels and, in particular, suggest how complementary value of information techniques might be applied to define the appropriate level of research and development.