Evaluating the longer-term sustainability of government programs and policies seems in many ways to go beyond the boundaries of typical evaluation practice. Not only have intervention failures over time been difficult to predict, but the question of sustainability itself tends to fall outside current evaluation thinking, timing and functions. This article is an effort to find some unifying concept that would allow evaluators to encompass the issue of sustainability within an evaluation framework, and to develop a set of criteria that could measure and account for unpredicted--but not necessarily unpredictable--changes in policy that have often affected sustainability in the past. What, for example, could evaluators do--without mingling in politics--to improve the quality of program or policy formulation, and hence potentially, sustainability itself? The author argues : (1) that there are (at least) 19 elements or values associated with the public interest in the United States--some consensual, some controversial--that can help clarify underlying policies, their likely changes, and their potential ramifications for sustainability; (2) that these values, and the preferred balances among them, would be different for different nations; (3) that some among these values will be more relevant than others to particular evaluations; (4) that there are both changes to current practice and new research that would eventually be needed; and (5) that both evaluation and sustainability could benefit from such an endeavor.