An improved methodology for quantitatively evaluating failure risk for a spaceflight system in order to assess flight readiness is presented. This methodology is of particular value when information relevant to failure prediction, including test experience and knowledge of parameters used in engineering analyses of failure phenomena, is limited. In this approach, engineering analysis models that characterize specific failure modes based on the physics and mechanics of the failure phenomena are used in a prescribed probabilistic structure to generate a failure probability distribution that is modified by test and flight experience in a Bayesian statistical procedure. The probabilistic structure and statistical methodology are generally applicable to any failure mode for which quantitative engineering analysis can be employed to characterize the failure phenomenon and are particularly well suited for use under the constraints on information availability that are typical of such spaceflight systems as the Space Shuttle and planetary spacecraft. Moore, N. R. and Ebbeler, D. H. and Creager, M. Jet Propulsion Laboratory NAS7-918...