In this thesis, the practice of map-making and map use is studied among actors involved in spatial planning and water management. The socio-technical mechanisms between knowlegde production and policymaking in Dutch regional planning make up the central object of study, with map images as observable artefacts. In many instances, maps seem to be magnets for conflict. Map images and their digital version embedded in geographic information systems (GIS) are generally described as helpful instruments that serve for supporting decision-making: but in many examples where multiple actors are involved, the policy maps that are used are surrounded by a lot of discussion. Connecting the literature fields of policy analysis and theory on (GIS-) cartography, Carton develops a framework of analysis with the concepts of frames and framing as central notions. With an argumentative research approach two cases are studied in-depth, to empirically analyze the functionality and effectiveness of maps from different actor perspectives. The first case reconstructs the making of a new long-term water policy in the region of water board Delfland. In the second case, a simulation game has been played with multiple actors of the province Brabant about a regional (urban/rural) spatial strategy. From the cases, three different frames of reference are identified, each with unique dominant values and inherent logic. These frames have either: - a scientific background where the map is considered a research model (analysis frame); - an attitude as creator or innovator where the map is considered to be a language to express one's ideas (design frame); - a political or negotiation attitude with corresponding assumptions and values where the map is considered a strategic agenda for making decisions (negotiation frame). It is argued that the differences between the three generalized frames 'analysis', 'design' and 'negotiation' explain many controversies over maps in the Netherlands that cannot be explained by 'simpler' explanations such as conflicting interests or information imbalance. Carton argues that this division of frames is part of the Dutch deliberative policymaking culture (referred to as "polder model"), and difficult to change. Furthermore, five strategies are identified how actors in practice cope with emerging map conflicts. On the basis of these findings, a number of recommendations are given for map makers/policy analysts in their work to support multi-actor policymaking.