DescriptionNavigating through an overhead visual maze is a demanding task. It relies on the strategic use of eye movements to select and identify the route. Maze solving also makes demands on memory and vision, and requires frequent decision and plans. When solving a maze, there are trade-offs between spending time exploring to the environment and spending time learning from errors. The current study examined strategies used to solve novel and familiar mazes that were viewed from above and traversed by a mouse cursor. Recorded eye and mouse movements revealed two modes that almost never occurred concurrently: exploration and guidance. Analyses showed that people learned mazes and were able to devise and carry out complex, multi-faceted strategies that traded-off visual exploration against active motor performance. The results challenge the previous findings that people prefer to use the external world as an external memory and minimize the use of the own short-term memory. Instead, people balanced the use of memory and the access to the external world, and this balance varied among different individuals. Overall, strategies of maze-solving took into account available visual information, memory, confidence, the estimated cost in time for exploration, and tolerance for error. Maze-solving provides an environment in which people have to continuously make decisions and plan paths in real time. By modeling the strategies people use, it is possible to draw inferences about many aspects of cognitive processes, such as the real-time decision making, the usage of memory in natural tasks and eye-hand coordination. The understanding of the strategies in maze solving may also benefit applications, such as designing navigation assistive devices and the development of methods to coordinate the interaction between human and machines (including robots) in road guidance.