The homing behavior of Lycosa tarentula (Linnaeus 1758) (Araneae, Lycosidae) adult females was studied. They were tested under two conditions, diffused light (200 lux) and darkness, after having been placed in an open field. In both conditions the spiders did not orient towards the burrow position; instead, under diffused light, each spider turned at a constant angle with a value close to135°; this is the turn that the spider should have made in its terrarium to return to the burrow. In darkness, most of the spiders (71.4%) turned at random. In both conditions, the trajectory was roughly straight, finishing with a sudden directional change. The speed was higher under diffused light than under darkness. These results support the hypothesis that L. tarentula uses path integration in laboratory conditions and that it needs visual input to obtain a direction estimation in homing.