The effects of postnatal depression on cognitive test scores at 20 months and 4;8 years of age as well as the timing (onset in the early postnatal period versus later), severity, number of episodes, duration of longest phase, recency, and chronicity of material depression on children's cognitive scores at 6;3 years was investigated. In South Bavaria, Germany, 1329 mothers of singletons were screened when the children were 6;3 years of age for the presence of depressive symptoms since the birth of their infant. A standard interview (SADS-L) was used to ascertain DSM-IV diagnosis and details of depressive episodes. Ninety-two mothers were diagnosed as having suffered DSM-IV defined depression (7%). Seven hundred and twenty-one mothers had no depressive episodes or symptoms from their children's birth until 6;3 years and were used as control group. The children had been assessed with the Griffiths Scales of Babies' Abilities (20 months), the Columbia Mental Maturity Scales (CMM) at 4;8 years, and the Kaufman Assessment Battery for Children (K-ABC) at 6;3 years. No significant main effects of severity, timing of onset, duration, or chronicity of depression of the child's cognitive development were found. Significant interactions of gender with chronicity of maternal depression (i.e. early-onset major and repeated episodes) were detected. Low SES boys or boys born at neonatal risk of mothers with chronic depression had lower Achievement Scores in the K-ABC at 6;3 years than children of mothers with less severe depression or controls. It is concluded that maternal depression per se has negligible effects on children's cognitive development. Long-term effects may be found when maternal depression is chronic, the child is a boy and neonatal risk-born, or the family suffers other social risks.