Growth of interdisciplinary research and education is accompanied by uncertainty about how to evaluate interdisciplinary work. What constitute indicators of quality that distinguish the exemplary from the mundane? What research evaluation processes are most appropriate when disciplinary standards do not suffice? To advance the discussion of quality assessment, researchers from Harvard University and the American Association for the Advancement of Science convened a select group of leading research administration (eg National Institutes of Health, Duke University), science journal editors (eg Physics Review) and social scientists, to share innovative practices and empirical results. Core insights are here summarized.