Prosocial behaviour is a bedrock of humane societies. Although previous studies have shown that parental style is an important contributing factor for children's prosocial behaviour, the psychological mechanism underlying this association in specific cultural background remains unclear. Based on the ecological systems theory, this study sought to explore the influence of parental warmth on children's prosocial behaviour, and the mediating effect of self-efficacy, as well as the moderating role of teachers' incentive evaluation in the link between self-efficacy and children's prosocial behaviour. This cross-sectional, survey-based study collected data from parents and teachers of 414 Chinese preschoolers (5–6 years old). Parents of these children finished questionnaires on the Chinese version of the short-form Egna minnen av. Barndoms uppfostran (s-EMBU-C), Child Behaviour Scale (CBS), and the revised General Self-efficacy Scale (GSES), while Incentive evaluation questionnaire was finished by their teachers. Results indicated that children who experience high parental warmth were more likely to act prosocially (b = 0.61, p < 0.001) and the association between parental warmth and prosocial behaviour was partially mediated by self-efficacy (indirect effect = 0.25, 95 % CI = [0.19, 0.32]). Besides, teachers' incentive evaluation moderated the link between self-efficacy and children's prosocial behaviour (b = 0.24, p < 0.001). Specifically, the effect of self-efficacy on prosocial behaviour was stronger for high teachers' incentive evaluation children than those with low teachers' incentive evaluation. These findings extend the existing understanding of the mechanism concerning the influence of parental warmth on prosocial behaviour in Chinese settings. The results revealed that interventions that could improve children's self-efficacy would be effective in accelerating their prosocial behaviour and schools played a crucial role in working with families to increase their self-efficacy at a young age. It should be noted that this study focused only on Chinese children of 5–6 years old, and the cross-sectional design means that the possibility of bidirectional effects could not be ruled out.