Nature of rod-touching behavior of cats was studied in two experiments. In Experiment 1, one of the four cats showed frequent flank/head rubbing behavior to the slender vertical rod in the cage when a human observer was present but it did not when the observer she was absent. The second cat “played” with the rod with its forepaws but rubbing behavior occurred mostly to the front net-wall of the cage. Remaining two cats sat immobile in a rear corner of the cage. In Experiment 2, it was observed that the first two cats tended to move their positions in accordance with the change of positions of the observer, thus keeping the closest distance with the observer. The last two cats again were little affected by the observer's position. The present study then showed that cats' pole-pressing behavior, interpreted by Guthrie & Horton (1946) as instrumental behavior, was an innate species-typical behavior of cats elicited by the presence of human observer and that the affectionate relationship of cat with the observer was one of the important determinants for the appearance of pole-pressing behavior.