Objectives: An increasing number of studies have presented evidence that various patient groups with acquired brain injury suffer from navigation problems in daily life. This skill is, however, scarcely addressed in current clinical neuropsychological practice and suitable diagnostic instruments are lacking. Real-world navigation tests are limited by geographical location and associated with practical constraints. It was, therefore, investigated whether virtual navigation might serve as a useful alternative. Methods: To investigate the convergent validity of virtual navigation testing, performance on the Virtual Tübingen test was compared to that on an analogous real-world navigation test in 68 chronic stroke patients. The same eight subtasks, addressing route and survey knowledge aspects, were assessed in both tests. In addition, navigation performance of stroke patients was compared to that of 44 healthy controls. Results: A correlation analysis showed moderate overlap (r=.535) between composite scores of overall real-world and virtual navigation performance in stroke patients. Route knowledge composite scores correlated somewhat stronger (r=.523) than survey knowledge composite scores (r=.442). When comparing group performances, patients obtained lower scores than controls on seven subtasks. Whereas the real-world test was found to be easier than its virtual counterpart, no significant interaction-effects were found between group and environment. Conclusions: Given moderate overlap of the total scores between the two navigation tests, we conclude that virtual testing of navigation ability is a valid alternative to navigation tests that rely on real-world route exposure. (JINS, 2016, 22, 467–477)