The metacognitive model has increased our understanding of the development and maintenance of generalized anxiety disorders in adults. It states that the combination of positive and negative beliefs about worry creates and sustains anxiety. A recent review argues that the model can be applied to children, but empirical support is lacking. The aim of the 2 presented studies was to explore the applicability of the model in a childhood sample. The first study employed a Danish community sample of youth (n = 587) ages 7 to 17 and investigated the relationship between metacognitions, worry and anxiety. Two multiple regression analyses were performed using worry and metacognitive processes as outcome variables. The second study sampled Danish children ages 7 to 12, and compared the metacognitions of children with a GAD diagnosis (n = 22) to children with a non-GAD anxiety diagnosis (n = 19) and nonanxious children (n = 14). In Study 1, metacognitive processes accounted for an additional 14% of the variance in worry, beyond age, gender, and anxiety, and an extra 11% of the variance in anxiety beyond age, gender, and worry. The Negative Beliefs about Worry scale emerged as the strongest predictor of worry and a stronger predictor of anxiety than the other metacognitive processes and age. In Study 2, children with GAD have significantly higher levels of deleterious metacognitions than anxious children without GAD and nonanxious children. The results offer partial support for the downward extension of the metacognitive model of generalized anxiety disorders to children.