Whilst considerable attention has been bestowed upon social competencies within the adolescent years in Germany (the Constance Youth Study, the BIJU-Study, PISA 2000), primary school age still remains to be studied with regard to social skills. The international school achievement study TIMSS 2007 (Trends in International Mathematics and Science Study) replies to this scientific desideratum by implementing facets of social competence into the German national questionnaire for the fourth graders as well as into the respective questionnaires for their parents and teachers. The presentation will set out delineating aspects of instrument development and focus on analyses with the resulting national TIMSS data set which is representative for fourth graders in Germany. Analyses will prioritize an evaluation of the instrument by emphasizing aspects of instrument reliability and validity. Furthermore, the empirical relationship between reports on social skill and (school) achievement outcomes – achievement on the TIMSS test, grades and intelligence – will be discussed. In this respect it will be of interest, in what way this relationship differs according to the external assessor on the one hand (the parent or the teacher) and to the measure of cognitive outcome on the other hand (achievement on the TIMSS test, grades or intelligence).