This study quantifies the dynamics of maternal mood focusing on unpredictability, and to assess if greater unpredictability of prenatal maternal mood predicts child temperament and internalizing symptoms through early adolescence.The association between prenatal mood predictability and child internalizing symptoms were assessed in two longitudinal cohorts (N's = 227 and 180). Maternal mood was assessed repeatedly during pregnancy as early as 15 weeks' gestation. Predictability of maternal mood was calculated by applying Shannon's entropy to the distribution of responses on mood questionnaires. Maternal reports of child negative affectivity (a predictor of later internalizing) were collected at 6, 12, 24 months and 7 years of age. Child self-reports of anxiety symptoms were collected at 10 years and reports of depression symptoms at 13 years.Fetal exposure to more elevated maternal mood entropy predicted higher levels of child negative affectivity at 12 months (r = .36; p < 01), 24 months (r = .31; p < 01) and 7 years (r = .32; p < 01) of age. In addition, children exposed to higher prenatal maternal mood entropy, reported higher levels of anxiety symptoms at 10 years (r = .24; p < 01) and elevated depressive symptoms at 13 years (r = .29; p < .01). These associations persisted after adjusting for maternal pre and postnatal mood valence (e.g. depression levels) and for other relevant demographic characteristics.Our findings provide strong support for the notion that patterns of maternal mood influence the developing brain. More specifically, they suggest that prenatal maternal mood predictability may be a critical predictor of developmental mental health trajectories and should be considered when assessing early life influences on lifespan mental health.