This article proposes a research program with two goals: (a) to support nonprofit leaders to productively engage evaluation and (b) to advance a meso-level theory of nonprofit evaluation that recognizes the diverse ways nonprofits contribute to social change. Such a research program is timely, as evaluation becomes increasingly institutionalized in the sector in ways that constrain nonprofit leaders from engaging productively with evaluation to advance their social impact. This research program brings existing nonprofit scholarship into conversation with evaluation scholarship and puts forward a research agenda organized around the practical dilemmas facing nonprofit leaders as they answer four key evaluation questions: what to evaluate, for what purpose, using which criteria, and with what evidence and methods. By anchoring a research program around these four questions, we seek to reopen the possibilities for how scholars can support nonprofit leaders in engaging evaluation to enhance their social impact.