The goal of this study was to predict the development of anxiety in 21/2 year olds as a function of maternal anxiety and child inhibited temperament, and to test the mediating, moderating, and curvilinear effects of maternal sensitivity. Participants were 83 mothers and their 21/2-year-old children (32 females). Maternal anxiety, child inhibition, and child anxiety were assessed by maternal report. Maternal sensitivity was rated based on the appropriateness and timeliness of mothers' responses to children's fear observed during their exposure to novel events in the laboratory and from mothers' diaries documenting their responses to children's fear in everyday situations. Gender predicted child anxiety, with mothers reporting girls as more anxious, as did child inhibition, with more inhibited children exhibiting more anxiety. Maternal sensitivity predicted child anxiety as a main effect and, in addition, inhibition moderated the curvilinear association of maternal sensitivity and child anxiety. For highly inhibited children, maternal sensitivity predicted anxiety in both a negative linear and a curvilinear fashion; anxiety decreased as maternal sensitivity increased up to a moderately high level, then increased at very high levels of maternal sensitivity. For less inhibited children, maternal sensitivity showed only a significant negative linear association with child anxiety. © 2010.