BACKGROUND: Screening instruments are often used for detecting mental health problems in children and adolescents. The Strengths and Difficulties Questionnaire (SDQ) is one instrument for screening children’s mental health. The SDQ can be used for assessment by different informants, i.e. parents, teachers and by 11-16 year olds for self-reporting. AIMS: The aim was to compare the precision and validity of parental and teacher SDQ assessments in elementary school children, and to analyze whether assessments were affected by the child’s sex and by socio-demographic factors. METHODS: A total of 512 primary school students were included in a cross-sectional study. Exploratory factor analysis, sensitivity/specificity analysis, Cronbach’s alphas, and logistic regression were applied. RESULTS: Parents rated 10.9% and teachers 8.8% of the children as high-risk individuals, but the overlap was low (32.1%). Cronbach’s alphas were 0.73 and 0.71 for parents and teachers, respectively. However, factor analysis showed that the five-factor solution could be confirmed only for teacher ratings. Moreover, only the parents’ ratings were affected by maternal educational level and parental country of birth when rating the same children as the teachers. CONCLUSIONS: Construct validity was only confirmed for teacher assessments. However, parental assessments might capture a dimension of a child’s mental health that seems to be sensitive to socioeconomic factors, which could be important when addressing equity issues, and for the dialogue between parents and school.