The purpose of the present action research is to examine the effects of three types of persuasion-skills on improvement in attitude and behavior of elementary school children. The subjects were two hundred and fourteen fifth and sixth graders. They were divided into four groups, The one control and three experimental groups. The communicator, who was an ostensible elementary school teacher, presented all four groups a standard persuasive message that they should not call other children by nicknames which annoyed them. In addition to this message, each of the three experimental groups was given a different experimental manipulation. These three were: (a) presenting an episode in which a child had negative experience as a result of being called by his nickname, (b) showing acceptance and sympathy with the nicknamer's emotional background, and (c) requesting to commit not to calling others by nicknames. All of three experimental groups showed greater improvement in attitude and behavior than the control group.