Description: Research on teacher perceptions has usually focused on judgments of student achievement. However, teacher perceptions do not only relate to the children's purely academic characteristics. Conceptions also have to do with students' social and emotional characteristics. Those conceptions are not necessarily explicit, but can presumably influence teachers' judgments of students' academic abilities and achievement in a significant way, and by this means affect their expectancies and assessment of students. Although the question of the effects of teacher expectancies for academic achievement is rather controversial, implications of such perceptions (referring to students' social, emotional and/or cognitive/academic characteristics) for student assessment can hardly be questioned. In the following research, we intend to highlight and analyse "contamination" effects between those different spheres of teacher perception in particular through the analysis of the link between academic expectations and teacher perception of children's social acceptance. Methodology: More than 4000 2nd and 3rd grade primary school pupils participating in a national longitudinal study of grade retention (founded by the Swiss national science foundation, see Bless, Bonvin & Schuepbach, 2005) were selected for the present study. Teacher evaluations of student characteristics were obtained through ad hoc scales (from the Teacher-Student-Questionnaire: Bonvin, 2003) for the following variables: intellectual potential, maturity, aggression, disruptive behaviours, social acceptance, and direct expectations for student achievement on both a mathematics and language test. On the student level, academic performance was measured with the above mentioned tests (on the basis of 2nd and 3rd grade curricula, see Bonvin, 2003). Measures further involved a sociometric analysis based on Krueger's approach (Krueger, 1976). Design is transversal, and results are yielded from correlational analysis, and a regression model of effects is presented. Conclusions: Preliminary results show the overall accuracy of teacher perceptions of student characteristics, especially when direct expectations are measured (here, in the case of academic achievement). Main results reveal significant correlations, on one hand between teachers' perception of students' behaviour (aggression, disruptive behaviours) and social acceptance (as perceived by the teacher) and, on the other hand, between perceived social acceptance and teacher expectancies for academic achievement. Those relationships between teacher perceptions (about students' academic and social characteristics) are found to a much lesser extent between objective measures of popularity (sociometric scores) and academic achievement as measured on independently evaluated achievement tests. In other words, teacher expectations for student achievement in academic areas are "contaminated" by a certain number of characteristics attributed to the student, which are not directly relevant to academic achievement.