Many current water resources management and planning problems are riddled with high levels of complexity, uncertainty and conflict; so-called unstructured (Kolkman et al. 2005) or messy (Ackoff, 1974) problems. These problems, which used to remain the domain of technical water managers, are increasingly entering the public policy sphere as conflicts between water users and interest groups proliferate and add a previously over-looked social dimension. The realisation that there is a need to consider the values and preferences of these stakeholders in decision making processes related to water management and planning has led to a multitude of new methods and processes being proposed for this process (HarmoniCOP, 2005), one of which includes participatory modelling (Hare et al., 2003). Unlike traditional modelling carried out by one person or institution, which may or may not include information from other stakeholders and which is used for as a decision aid, participatory modelling allows a number of different points of view to be explicitly represented and collectively reflected upon by a group of stakeholders before a collective decision is made (Ferrand, 1997). Despite an increasing uptake of such methods, little research and investigation of the concrete advantages and disadvantages of using these methods has been undertaken. This paper stems from the preliminary phases of a research project which aims to investigate the utility of a participatory modelling process for water management and planning applications, such as improving collective water use and fostering the adoption of more sustainable practices and/or technologies. More specifically, the objectives of the research are to: (i) determine the positive and negative aspects of using a participatory modelling approach for decision aiding in water resources planning and management applications (descriptive objective); and (ii) determine under which circumstances a participatory modelling approach is likely to be more effective than externally produced models for the achievement of sustainable water management (prescriptive objective). Related to these objectives, the hypotheses to be tested include that participatory modelling: (1) helps to examine the real underlying problems; (2) increases trust, appropriation and understanding of the models created as assumptions and uncertainties are more likely to be explicitly identified and discussed; (3) generates greater creativity and innovation; (4) leads to an improved ability to respond to change through enhancing social capacity, adaptability, flexibility and resilience; (5) leads to greater individual and social learning; (6) produces richer and more realistic action plans; and (7) provides a greater chance of adoption or implementation of problem solutions. This paper specifically aims to outline only the first phases of this research project which include: a brief review of current understanding related to participatory modelling and the development of a participatory modelling classification system, as well as a couple of best practice participatory modelling methods. It will also briefly present a new participatory modelling methodology which has been developed based on this review and is currently in the testing and validation phase. After improvements, this methodology is planned to be used as a support for water planning and management applications to help the investigation of the above hypotheses. The methodology takes the form of six consecutive phases (likely to involve feedback loops) in which all stakeholders or participants will participate and continuous evaluation will occur: 1.Introduction and problem situation / 2.Problem formulation / 3.Model construction / 4.Model testing and construction of rules for role analysis / 5.Scenario exploration through role analysis / 6.Decision or choice and implementation of options. The design of the methodology attempts to bring together decision-aiding theories (i.e. Tsoukiàs (2005)), planning cycles (i.e. Holling (1978), Deming (1986) and Strömgren (2003)) and important aspects of currently available participatory modelling protocols (i.e. Palmer et al. (1993), Barreteau (2003a), van den Belt, (2004)). Testing and validation of the methodology is currently at an intermediary stage, following a recent trial with a group of tertiary students in Montpellier, France, who examined their own water related issues at multiple (although abstract) spatial scales (their life and home, their neighbourhood or village, and their water basin or region). The trial took place in the form of seven consecutive workshops over a period of a month and a half and used a variety of methods throughout the linked phases, including cognitive mapping, an adapted version of Ackermann and Eden's (2001) Oval Mapping Technique, UML model design and the construction and use of a role playing game. Evaluation procedures (which included 15 questionnaires and collective reflection and feedback periods) also constituted approximately twenty percent of the total process time. The utility of such a test case in terms of validating a methodology for real-life multi-stakeholder water management and planning processes must obviously be questioned, but for institutional reasons it was preferred due to its minimal risk. Some specific methodological choices made because of the participating group, such as paying participants for their participation costs or using a role playing game that was specifically chosen to give the students a forum and mock situation for collective decision making, are not likely to be used in other real cases. Despite this drawbacks, preliminary results from the test have provided a wealth of interesting information and have helped to uncover many useful questions surrounding: model complexity (which stems from the examination of complex problems); process complexity; if, how or when to include exterior information (from a facilitator's or outside expert's point of view); time management; and the importance and cost of the evaluation process. Other results of the process have included positive feedback in terms of the educational value and creativity that occurred in the problem situation and problem formulation phases. However, at this stage, for the methodology to be used for more practical purposes, it still requires further adjustment. In particular, stakeholders to be involved in a participatory modelling process applied to real water planning or management problems (those problems which exhibit a good degree of stakeholder disagreement, uncertainty and complexity) should be involved in the redefinition of the process and the methods used throughout this process