Theory and preliminary evidence suggests that parental beliefs and cognitions may be transmitted to their offspring. Transmission of maladaptive cognitions may play a role in the development of anxiety disorders in childhood. However, few studies have investigated such transmission using longitudinal designs. The objective of this study was to investigate the interaction between maternal and child metacognitions and their role in the development of childhood anxiety. We used a longitudinal design with self-report measures of maternal and child anxiety symptoms and metacognitions. Participants were 107 mothers and their children who were assessed when the children were between 7 and 12 years old and again 3 years later. Child metacognitions at baseline did not predict later child anxiety symptoms. Baseline maternal metacognitions approached significance in predicting anxiety symptoms in children at the follow-up, when controlling for known risk factors, including female gender and higher levels of anxiety in mother and child at baseline. Mediation analyses revealed that child metacognitions at baseline, as well as at follow-up fully mediated the relationship between maternal baseline metacognitions and child anxiety 3 years later. Examinations of how child and maternal metacognitions affect child anxiety levels after three years revealed that current levels of child metacognitions play a greater role in predicting child anxiety than child and maternal metacognitions at baseline.