Freshwater resources are coming under increasing pressure across New Zealand. High levels of water use and contaminant loss from both point and non-point sources are contributing to growing concern around both the quality of freshwater and its availability. National government regulation has provided greater guidance on the freshwater standards required in water ways, while increasingly collaborative processes are being applied to develop directions for both national and regional policy. A shift towards the inclusion of stakeholders in policy development follows global trends; however, its application in reality is difficult, due to scarce resources (especially data and time) and the high biophysical and social complexity of the issues being addressed. This research investigates whether the use of causal-loop diagrams (CLDs), applied in a group model building (GMB) process, increases the shared understanding of a freshwater issue by a group. CLDs are a qualitative system dynamics (SD) modelling tool, which seeks to identify the interdependency between elements of a system and the direction of these relationships. They are straightforward to generate and can provide insight into the underlying causes of suboptimal behaviour in complex, dynamic environments. The GMB process applied in this research involves the extension of an existing script-driven process, after extensive piloting. It is applied with only two supporting personnel within a limited timeframe, to explore how useful this process is within an environment involving scarce resources. The developed process is applied to two case studies, both involving freshwater management by New Zealand communities. The effectiveness of the GMB process was explored using a mixed methods approach. A post-workshop survey gathered self-reported quantitative and qualitative findings, while a qualitative analysis of data from semi-structured follow-up interviews was undertaken using thematic analysis. The self-reported data indicated not only very high levels of individual satisfaction with the process, but also that the process had helped the group’s shared understanding of the issues and that the output (the CLD) would be useful to the group in the future. The thematic analysis identified four broad themes: that there was a desire for a different approach to decision-making; that the GMB process was a refreshing change; that the GMB process built both a group AND a group model; and that, while it was useful, it also did not provide enough detail for decision-making on its own. The desire for a different approach was underpinned by the complexity of the decision-making process; as well as frustration from some of those involved that they either did not feel that their views were being (or had been) appreciated or from a direct sense of frustration from dealing with council (currently or in the past). It was felt that the GMB approach was a refreshing change as it was a simple and focused process, involving a small group of relevant stakeholders. More importantly, the fact that the group members were actively involved in the process, it was participant led and independently facilitated were all very positive aspects identified. However, the process for GMB that was applied was found not to be a panacea for solving problems, with more detailed, numerical modelling required to provide insight into the relative importance of key factors.