Introduction In this paper we address the steps that precede the creation of a mathematical representation of real world phenomena. More specifically, we investigate first the formulation of the ‘problem statement’ in an integrated assessment modelling study, seeking here to offer a more precise definition of problem formulation and examining its influence on later stages of modelling. We identify, through both numerous group discussions at the summer school and a survey of the Integrated Assessment literature, that although there is much discourse around the ‘problem’ formulation, it has not been well defined. Subsequently, a more precise definition may aid in helping us to understand the steps researchers might take in formulating and solving a problem. Furthermore, problem formulation will be improved through additional classification and detailed analysis particularly with respect to stakeholder involvement and participatory approaches. These steps may be of value to not only emerging researchers and professionals such as has attended this current and future Summer Schools, but also giving credence to the observed lack of attention to problem formulation in the literature.