We examined the association of prenatal depressive symptoms at mid-pregnancy with child cognition at age 3 years in Project Viva, a pre-birth cohort study of 1030 mother-child pairs in eastern Massachusetts. We measured maternal depressive symptoms using the Edinburgh Postnatal Depression Scale (EPDS), a self-report measure validated for use during pregnancy. Measures of child cognition included the Peabody Picture Vocabulary Test (PPVT) and the Wide Range Achievement of Visual Motor Abilities (WRAVMA). At mid-pregnancy, 81 mothers (7.9%) scored 13 or above on the EPDS, indicating probable depression. In the unadjusted model, children born to mothers with prenatal depressive symptoms had PPVT scores that were 3.8 points lower [95% confidence interval (CI) -7.1, -0.5]. With adjustment for sociodemographic variables, the association substantially attenuated [adjusted regression coefficient b for PPVT score = -0.7 (95% CI -3.6, 2.3)]. In both unadjusted and multivariable models, prenatal depressive symptoms were not associated with WRAVMA scores [adjusted b for total WRAVMA score = -0.5 (95% CI -3.0, 2.1)]. We found no evidence to suggest that maternal prenatal depression is independently associated with early child cognition.