This study examined the interplay between maternal depressive symptoms and emotion coaching in predicting preschoolers’ emotion regulation skills. Participants included 126 preschoolers and their mothers, assessed at two time points, when children were 3 (T1) and 4 (T2). Mothers’ coaching of sadness and anger was observationally assessed in a laboratory task at T1 under two categories, high and low elaborative styles of emotion coaching. Mothers also reported their depressive symptoms at T1. Children’s emotion regulation was observed in laboratory tasks eliciting sadness and anger at T1 and T2. Path analyses revealed three-way interactions among maternal depressive symptoms, low elaborative emotion coaching, and high elaborative emotion coaching in predicting preschoolers’ emotion regulation in both sadness- and anger-eliciting tasks 1 year later. Maternal low elaborative emotion coaching of sadness reduced children’s focus on distress in the sadness-eliciting task only when mothers showed high depressive symptoms and used little high elaborative emotion coaching. In contrast, maternal low elaborative emotion coaching of sadness and anger predicted maladaptive child emotion regulation in both sadness- and anger-eliciting tasks, respectively, when mothers were depressed or used greater high elaborative emotion coaching. Findings highlight the importance of considering different types of maternal emotion coaching language styles in the context of maternal depression and have implications for intervention efforts.