Identifying environmental risk factors in the pathway to depression is an important research goal. To extend prior research, designs that rule out alternative explanatory factors; genetic effects and reverse causation, and permit tests of both parent and child gender are required. The present study used two different samples to address these issues. A longitudinal community sample of 316 families (157 boys, 159 girls) aged 11-12 years (mean 11.7) at Time 1 and 12-13 years at Time 2 (mean 12.7) was used to test the direction of effects between parent hostility and child and adolescent depression symptoms. A genetically sensitive sample of 1,075 twin pairs; 653 dizygotic (135 male, 183 female, 335 opposite sex) and 422 monozygotic (180 male and 242 female) aged 12-20 years (mean 16.12) was used to test whether parent hostility had environmental effects. Analyses were conducted separately by parent and child gender. Using cross-lagged panel analyses, the association between mother-daughter hostility and depression symptoms was found to be longitudinal and bidirectional with reciprocal effects between mothers and daughters. Behavioural genetic analyses in the twin sample revealed a significant environmental link between mother hostility and symptoms of daughter depression independent of genetic factors. A significant pathway was found between daughter depressive symptoms and father hostility but not vice versa. This association was accounted for by genetic factors in behavioural genetic analyses. Findings provide evidence of an environmental risk pathway to depression symptoms and identify patterns of variation according to parent and child gender. Results are discussed in relation to underlying explanatory processes and clinical implications.