Heritability estimates for adolescent anxiety vary across studies, partly depending on who is rating the symptoms. The goal of our study was to estimate genetic and environmental influences using a multi-informant design with responses from a population-based sample of adolescent twins, their mothers and their fathers (N=1394 families). Results from multivariate biometrical modeling indicated quantitative, but no qualitative sex differences in etiology. The best fitting model was an AE Common Pathway model, defining anxiety as a latent factor common to all informants. This model offers error free estimates of genetic and environmental influences explaining the latent factor variance. Variation in the latent factor was highly genetic, with heritability estimates of 65% for boys and 74% for girls. Non-shared environmental effects explained the remaining variance. In addition, there were significant rater-specific genetic and environmental effects for both sexes. The observed rater differences underline the importance of using several informants when studying adolescent anxiety.