Spatial developments are becoming more and more non-linear, dynamic and complex with a wide range of possible actors. The awareness of uncertainty in spatial planning is growing and therefore, projects need to integrate a high level of flexibility. But at the same time, a growing demand for taking more informed and well-argued decisions is noticeable. Predictions out of the ‘best estimated model’ are no longer credible and no longer accepted, because they are too fragile and uncertain. How can we keep these long-lasting, multi-actor projects in permanent transition on the right track? This article presents an evaluation methodology that goes beyond the traditional, rational evaluation attitudes with a low level of flexibility being too linear to match the current spatial developments. There is a need for more interrelated, alert and flexible means of evaluation, co-evolving with the processes and current dynamics in spatial planning. Therefore, different evaluation approaches are introduced, depending on the specific interdependencies of the object of evaluation and its context. Subsequently, the theoretical framework is translated towards a more practical level. A case study conducted in Flanders illustrates the current spatial developments and a possible evaluation approach, incorporated from the beginning of the process, to guide this kind of projects.