The big brown bat, Eptesicus fuscus, uses echolocation for foraging and orientation. The limited operating range of biosonar implies that bats must rely upon spatial memory in familiar spaces with dimensions larger than a few meters. Prior experiments with bats flying in obstacle arrays have revealed differences in flight and acoustic emission patterns depending on the density and spatial extent of the obstacles. Using the same method, combined with acoustic microphone array tracking, we flew big brown bats in an obstacle array that varied in density and distribution in different locations in the flight room. In the initial experiment, six bats learned individually stereotyped flight patterns as they became familiar with the space. After the first day, the repetition rate of sonar broadcasts dropped to a stable level, consistent with low-density clutter. In a second experiment, after acquiring their stable paths, each bat was released from each of two unfamiliar locations in the room. Each bat still followed the same flight path it learned originally. In a third experiment, performed 1 month after the first two experiments, three of the bats were re-flown in the same configuration of obstacles; these three resumed flying in their accustomed path. The other three bats were flown in a mirror-image reconfiguration of the obstacles; these bats quickly found stable flight paths that differed from their originally learned paths. Overall, the flight patterns indicate that the bats perceive the cluttered space as a single scene through which they develop globally organized flight paths.