The puzzle of NATO's persistence is best addressed as part of a larger inquiry into institutional change. Institutions persist because they are costly to create and less costly to maintain, but this institutionalist argument is incomplete. Whether institutions adapt to change depends on whether their norms, rules, and procedures are specific or general assets and on whether the asset mix matches the kinds of security problems faced by their members. Assets specific to coping with external threats will not be useful for coping with problems of instability and mistrust, so alliances with only the former will disappear when threats disappear. Alliances that have specific institutional assets for dealing with instability and mistrust and general institutional assets will be adaptable to environments that lack threats. I assess these hypotheses in a test case of NATO's institutional assets during and after the Cold War.