Initiatives to map nonprofit organizations encompass efforts to define the boundaries of the sector and understand its scope and scale. As new technologies make it possible to digitize and analyze information in new ways, further questions about mapping civil society emerge. We integrate nonprofit scholarship, critical work on computational methods, and reflection on our experiences using machine learning to map nongovernmental organizations in Ghana, to develop a critical framework for mapping civil society in the digital age. The issues we raise about computational methods are embedded within greater concerns about the taken-for-granted assumptions in mapping civil society, and mapping as a tool to control, manage, and manipulate civil society. We are particularly attentive to the power within mapping as a mode of knowledge production.