[en] The objective of this study is to demonstrate how to specify containment performance criteria in a systematic way given top level safety goals. The multiobjective optimization approach proposed by Cho, et al. was adopted as a method for deriving a finite manageable set of self-consistent relations between the top level safety goals and specific sets of measures of containment performance. The current trial safety goals proposed by the NRC goes from a design objective figure of 10^-4 for annual frequency of core melt to guideline numbers for early fatalities and for delayed fatalities. An alternative is to specify containment system performance in some quantitative fashion. If this can be done, and if the specified performance can be demonstrated with some reasonable level of confidence, it will help to provide greater consistency between top level safety goals (health risk indices) and lower level design objectives (plant performance indices, i.e., core melt frequency and containment performance criteria)