<p>This study employs a longitudinal design and constructs a cross-lagged model to explore the dynamic interactions among family functioning, parent-child attachment, and depressive symptoms in Chinese children. The study involved 624 third-grade students from three primary schools in a northern Chinese province, who were assessed longitudinally three times within one academic year, with six-month intervals between each assessment. The results of the cross-lagged model indicate that: (1) when accounting for the stability of each variable over time, family functioning significantly predicts depressive symptoms via the mediating effect of parent-child attachment. Family functioning promotes a positive interaction process, where family functioning at the first measurement (T1) significantly enhances parent-child attachment at the second measurement (T2), thereby reducing the risk of depressive symptoms at the third measurement (T3). (2) Depressive symptoms drive a negative interaction process, where depressive symptoms at T1 significantly impair family functioning at T2, thereby worsening parent-child attachment at T3. The study results support the transactional model, indicating that good family functioning can mitigate the negative driving effects of depressive symptoms and intervene in depressive symptoms through the protective role of parent-child attachment.</p>