The present experiment discusses (1) the use of "cognitive maps" in rats by showing that they can orient themselves and their direction of movement at a given place in a maze and (2) the role of the hippocampus in spatial cognition by examining the effects of hippocampal lesions on rats performing maze tasks. The apparatus employed was a lattice maze consisting of eight alleys which cross orthogonally. The alleys were partitioned by a total of 24 apparently identical doors; 8 doors were locked and the others were unlocked. Fimbria-fornix-lesioned (FF) (n = 16) and control (n = 18) male Wistar rats were trained to run from one of four start boxes to the goal with a salient visual cue under open and closed conditions. Rats in the OPEN-Cont group successfully discriminated between locked and unlocked doors and were able to choose detour routes as well as the shortest route. This was regarded as evidence of their ability to form a cognitive map. Rats in the CLOSED-Cont group did not have intra- or extramaze cues but were also able to learn the task. It is proposed that they used proprioceptive cues to form cognitive maps, which enabled them to solve the problem. Rats in the FF group could not avoid locked doors because they were not able to form cognitive maps. However, they utilized the goal stimulus as a cue to solve the problem.