Abstract In this paper, four frequently cited approaches to future studies are criticised. We use examples mainly from the field of transport research. The first approach is the tendency to try to establish cyclic behaviour in socio–technical changes. The second is the view that transport and communication are positively correlated. The third is the so-called ‘hypothesis of constant travel time’, according to which, the average daily travel time of a population is more or less stable. The fourth is the alleged causal relationship between urban density and petrol use. The use of these approaches is criticised for a number of reasons, among others for oversimplifying the underlying mechanisms and for being too deterministic. In cases where drastic change is needed, current trends must be broken, but perhaps through measures other than those indicated by the above approaches. In other words, the cited approaches may overlook interesting opportunities and fail to urge necessary action. Backcasting is put forward as a more promising approach, especially for situations where great change is needed. However, it has been found in this study that backcasting and different forecasting approaches are complementary. The argument is that backcasting is mainly appropriate where current trends are leading towards an unfavourable state. Therefore, forecasting methods are necessary because they inform the backcaster when backcasting is required. Finally, the paper discusses the use of different models in planning, primarily in the context of their role in the path analyses of backcasting scenarios.