Maternal depression has consistently been found to be associated with child problem behaviours. In the current study, we examine the possibility of whether the association between maternal depression and child antisocial behaviour may be due to confounding. Data from waves 2, 5 and 7 of the Early Childhood Longitudinal Study‐Kindergarten Cohort were analysed. Our analysis revealed a significant association between maternal depression and child antisocial behaviour prior to matching. After successfully matching the mothers on 29 key covariates, one of the cross‐sectional associations and the longitudinal association between maternal depression and child antisocial behaviour was no longer statistically significant. However, two cross‐sectional associations between maternal depression and child antisocial behaviour remained statistically significant. The results of the study are discussed in relation to the importance of the timing of maternal depression within different developmental time periods.