Weeds pose severe threats to agricultural and natural landscapes worldwide. One major reason for the failure to effectively manage weeds at landscape scales is that current Best Management Practice guidelines, and research on how to improve such guidelines, focus too narrowly on property-level management decisions. Insufficiently considered are the aggregate effects of individual actions to determine landscape-scale outcomes, or whether there are collective practices that would improve weed management outcomes. Here, we frame landscape-scale weed management as a social dilemma, where trade-offs occur between individual and collective interests. We apply a transdisciplinary system approach—integrating the perspectives of ecologists, evolutionary biologists and agronomists into a social science theory of social dilemmas—to four landscape-scale weed management challenges: (i) achieving plant biosecurity, (ii) preventing weed seed contamination, (iii) maintaining herbicide susceptibility and (iv) sustainably using biological control. We describe how these four challenges exhibit characteristics of ‘public good problems’, wherein effective weed management requires the active contributions of multiple actors, while benefits are not restricted to these contributors. Adequate solutions to address these public good challenges often involve a subset of the eight design principles developed by Elinor Ostrom for ‘common pool social dilemmas’, together with design principles that reflect the public good nature of the problems. This paper is a call to action for scholars and practitioners to broaden our conceptualization and approaches to weed management problems. Such progress begins by evaluating the public good characteristics of specific weed management challenges and applying context-specific design principles to realize successful and sustainable weed management.