The purpose of this study was to obtain data in order to make a Children's Social Support Scale. In two pilot studies, basic data was gathered by open-end questionnaires concerning the content and sources of social support appreciated by elementary school children. The pilot study I was administered to 78, 6th graders, while pilot study II to 96, 4th graders and 172, 6th graders. After item analyses, 11 items were selected, followed by the construction of a scale assessing Children's Social Support (CSS). The results regarding the reliability and the validity of CSS were as fo llows: 1) Study II (administered to 451, 4th to 6th graders) showed an internal correlation of r =.96, after an adaption to the Spearman-Brown formula. Moreover, the test / retest reliability was a high r =.89, confirming the reliability of CSS. 2) Positive correlation with another social support scale (Hisada et al. 1989) was r =.87, whereas the correlation with a child depression scale (Sakurai 1989) proved negative r =-.34.