This article argues that Congress should authorize and fund a National Academies risk analysis of nuclear deterrence. It also explains how risk analysis is able to tease much more information out of the available data – which clearly does not yet involve a failure of nuclear deterrence – than might first appear possible.