The creation, maintenance, and advancement of a complex signage system, as found, for example, in hospitals or airports, require a number of different decisions that affect both, a sign’s legibility and intelligibility. After routes have been planned globally, for each individual sign board the designer must decide (1) which information should be displayed, (2) how the information should be arranged and (3) where in space the sign should be installed. In addition, before a sign is put up, it is advisable to assess whether an additional sign will effectively support a user’s spatial decision making or if it simply adds to visual clutter. In this process a number of methods are commonly used to identify the optimal signage for a particular location. These methods include, for example, the consideration of design guidelines, which are often based on knowledge of human perception and are sometimes regulated by law (fire codes). Additionally, knowledge from former projects is often transferred to the current one and over time best-practice solutions get established. In the majority of cases these methods lead to the desired results, although the effectiveness and efficiency are rarely evaluated with respect to the user’s needs and the sign’s utility for the spatial decision making process. We developed a method that combines behavioral, self-assessment, and eyetracking data in order to assess the effect of direction signs on passengers’ decision making in an airport context. Participants were presented either with the original or with a digitally edited photographic image of a scene that required them to make a spatial decision in order to reach a given target location. The combination of decision errors, decision time, self-assessed confidence and the allocation of attention, measured through eye-movements, allowed us to assess different designs and the effectiveness of a newly placed sign with respect to user behavior. Based on this we were able to recommend a particular design solution. In a further study we tested different design options in a virtual model of a planned extension of the building. This scenario enabled us to give evidence-based recommendation, even before the site was actually built. We consider the method as a supplement rather than a replacement of existing methods for the development and advancement of wayfinding systems with special emphasis on cognitive adequacy for humans.