Consensus building and other forms of collaborative planning are increasingly used for dealing with social and political fragmentation, shared power, and conflicting values. The authors contend that to evaluate this emergent set of practices, a new framework is required modeled on a view of self-organizing, complex adaptive systems rather than on a mechanical Newtonian world. Consensus building processes are not only about producing agreements and plans but also about experimentation, learning, change, and building shared meaning. This article, based on our empirical research and practice in a wide range of consensus building cases, proposes that consensus building processes be evaluated in the light of principles of complexity science and communicative rationality, which are both congruent with professional practice. It offers principles for evaluation and a set of process and outcome criteria.