Community schools have long been accepted as an institutional mechanism for intervening in the relationship between poverty, poor educational outcomes, and limited life chances. At a time when public services are being retracted, and disadvantaged places are being increasingly left to struggle, community schools are poised to become more important in offering a response to the needs of children, families, and communities in these places. Yet, despite their apparent promise, community schools remain badly under-conceptualized. As an international field, research on community schooling has rarely articulated or questioned how--by providing additional learning and leisure opportunities and personal and social supports--community schools might create a viable intervention in the relationship between poverty and poor outcomes. This paper explicitly addresses this significant challenge. Conceptualizing empirical findings emerging from a research-practice partnership, it identifies the core features of a new institutional design for community schools which can help to clarify their potential contribution to addressing disadvantage. Marking a considerable shift from a traditional design of simply adding new services to the school day, it argues that community schools will need to operate as social enterprises with networked governance arrangements, and to develop strategies which engage with children's social ecologies, and are risk-reducing and resilience-building within these. This, in turn, sets a new agenda for significantly advancing the field of community schooling by further defining--conceptually and empirically--the core elements of a new institutional design as identified here.