Although adolescents’ prosocial behavior is related to various positive outcomes, longitudinal research on its development and predictors is still sparse. This 6-wave longitudinal study investigated the development of prosocial behavior across adolescence, and examined longitudinal associations with perspective taking and empathic concern. Participants were 497 adolescents (M _{age t1 =}13.03 years, 43% girls) who reported on their prosocial behaviors, empathic concern, and perspective taking. The results revealed marked gender differences in the development of prosocial behavior. For boys, levels of prosocial behavior were stable until age 14, followed by an increase until age 17, and a slight decrease thereafter. For girls, prosocial behavior increased until age 16 years and then slightly decreased. Regarding longitudinal associations, empathic concern was consistently related to subsequent prosocial behavior. However, perspective taking was only indirectly related to prosocial behavior, via its effect on empathic concern. Tests of the direction of effects showed support for the notion that earlier prosocial behavior predicts subsequent empathy-related traits, but only for girls. The findings support cognitive-developmental and moral socialization theories of prosocial development and the primary role of moral emotions in predicting prosocial behaviors. Our findings inform strategies to foster prosocial behaviors by emphasizing moral emotions rather than moral cognitions during adolescence.

Empathy, which is the ability to feel concern for and to understand others’ feelings, is thought to develop in high quality relationships with parent and peers, but also to facilitate the quality of these relationships. While a wide literature has addressed this aspect, the heterogeneity of primary studies, in which different indicators of relationship quality (e.g., support, conflict) and empathy (i.e., affective and cognitive) have been examined, makes it difficult to draw conclusive answers. Therefore, it remained ambiguous how parent–child and peer relationship quality are associated with adolescents’ empathy. In order to increase the understanding of these associations, a multilevel meta-analysis was performed, which allowed for including multiple effect sizes from each study. By a systematic literate search, 70 eligible studies were found that provided 390 effect sizes from 75 independent samples. The results showed a small positive correlation between parent–child relationship quality and empathy, and a small-to-moderate positive correlation between peer relationship quality and empathy, which was significantly stronger than the correlation with parent–child relationship quality. Hence, the meta-analytic results indicate that adolescents with higher quality relationships, especially with peers, indeed tend to show more concern for and understanding of others’ emotions than adolescents with lower quality relationships. Moreover, the moderation analyses showed stronger correlations for the positive dimension of relationship quality than for the negative dimension, and stronger correlations for composite scores of affective and cognitive empathy than for separate scores of the empathy dimensions. However, no differences in correlations were found between the affective and cognitive empathy dimension, and no moderation effects were found for gender and age. Thus, this meta-analysis demonstrates robust positive associations between parent–child and peer relationship quality and empathy in adolescence, implying that good empathic abilities may be a protective factor for experiencing poor relationships.