Wayfinding difficulties in architecturally ambiguous environments can be overcome by orientating with given landmarks. However, it is not clear at which locations landmarks are most suitable to facilitate orientation. Our study addressed two questions: (1) Is subsequent wayfinding performance facilitated if participants freely place a number of landmark objects in a complex building? (2) Where do participants place the landmarks and which strategies guide effective landmark placement? First, participants were instructed to learn a number of goal locations in a virtual model of the Tate Gallery London. Then, participants in the experimental condition were instructed to place five unique landmarks in order to re-find the goals in a third phase; participants in a control condition could not place landmarks. Finally, participants were tested on their ability to find the goals again, with time and distance as main dependent variables. Results are discussed with respect to placement strategies and environmental properties.