This study investigated developmental changes in strategies to determine a social preference order, given a set of individual preference orders. Experiment I examined how a social preference (decision as a group) was determined when each member's preference order was known. Seventy-four 5th graders and 86 undergraduates participated in this experiment. Subjects were shown patterns of individual preference orders of A, B, and C given by members of a group of three. One example is given below. _??_ The task was to decide upon a social preference order of three alternatives A, B, and C based on such preference order patterns. Experiment II aimed at examining more closely how children and undergraduates would make such decisions. Subjects were asked to describe in detail how they had decided such orders under two conditions. In PN (Positive-Negative) condition, each member's preference concerning the three alternatives were expressed as “Most desirable”,“Neutral”, and “Most undesirable”. In O (Order) condition, alternatives were all thought to be desirable, in a first, second, and third order. Seventy-two 5th graders, ninety 8th graders and eighty-six undergraduates participated in the experiment. The main results were as follows. 1. Majority of decisions made by subjects were consistent when patterns of the individual preference order given by three members were the same. 2. From analyzing subjects' own descriptions of the strategies of making decision, four underlying principles emerged:(1) minority-oriented majority rule (2) concession-oriented principle (3) majority rule (4) local majority rule. 3. While most of 5th graders adopted the local majority rule, 8th graders and undergraduates adopted the variety of principles. 4. As for 8th graders and undergraduates, the concession- oriented principle was adopted more often in the PN condition than in the O condition. Implications of these findings were discussed.