The purpose of this study was to examine the effect of different instructional methods adopted to individual differences on spatial learning in preschool children. The experiment consisted of three sessions; a pre-test, an experimental education session, and a post-test. In the experimental education session, the children's ability to orientate objects correctly in large scale environment was studied. In the pretest and the post-test, the subjects were asked to make a drawing of the disposition of their room at home and to draw a map showing the way from their kindergarten to the park located near by. In the post-test, they were asked to compose the map of the surroundings of their kindergarten, using the given seals that showed a road, a park, a building, and so on. 32 chi dren (5-year-old) were divided into two groups (16 subjects each) on the basis of the pretest score and contents. The subjects of one group were considered to have the reference system of their own body considered as a reference point to orientate objects (FP-group), and the subjects of the other group were considered to have slight indications of getting the reference system of “coordinated” orientation (FW-group). In order to teach those two kinds of subjects, two different instructional methods were made. In one method children were helped to learn how to use their own reference system usefully in order to orientate objects (B-type). In the other method childrenwere helped to obtain the reference system of geographic orientation as a reference point, which seemed to be a new idea to them (G-type). The B-type method was supposed to have effects on the FP-group and the G-type method on the FW-group. In the experimental education session, the FP-group and the FW-group were also divided into two groups, according to the given instructional method. Then four groups were made (8 subjects each): FP-B-group, FP-G-group, FW-B-group and FW-G-group. The main hypotheses were as follows; The FP-B- group and the FW-G-group would show higher score than the FP-G-group and the FW-B-group in the post-test. The FP-G-group and the FW-B-group would not improve their space cognition for lack of suitable instructional treatment. The results supported the ab ove hypotheses while indicating that the G-type method was effective for only the FW-G-group that received good marks in the pre-test, while the B-type method was effective for almost all the subjects of the FP-B-group. It was also shown that the score of the FP-group in the post-test was relatively lower than that of the FW-group. It was suggested that the B-type method was effective for the subjects of the FP-group so that they might improve their space cognition up to a certain level, but for a greater improvement, it would be necessary to give them different types of instructional method, which might be something of the G-type, in a following teaching process.