The paper deals with one of the oldest and still most pressing issue in system dynamics: “when to map and when to model”. We propose a heuristic to be used in system dynamics projects to give an indication whether one should aim at quantitative modeling (i.e., the classical system dynamics approach) or whether qualitative treatment will be sufficient: (1) If the problem definition phrase starts with the word “why”, often causal explanations for the development of variables should be the result of the project; in this case, the exploration of dynamic behaviour needs the application of quantitative simulation models because of the difficulties people have with inferring dynamics from complex models; (2) If the problem definition phrase starts with the word “how”, the structure of a system is asked for; then, the result of the project is usually a graphical representation of the shared understanding of team members about the structure of a system, which can be achieved by using qualitative modelling techniques like causal-loop diagrams, hexagons, or policy structure diagrams.