Given the heightened difficulties in social adjustment and the potential diminishment of social networks encountered by migrant children, family functioning may play a crucial role in their development. Existing research has highlighted the significance of family environment in shaping adolescent self-compassion and emotion regulation, which can serve as protective factors against adverse emotional outcomes. However, there remains a lack of comparative studies to examine the specific effects of family functioning on fostering self-compassion and emotion regulation in both migrant and their non-migrant counterparts. The present study utilized a three-wave longitudinal design with 12-month intervals to examine the longitudinal effects of family functioning on self-compassion and emotion regulation, while also examining potential variations in these associations between migrant and non-migrant children. A total of 244 migrant children and 491 non-migrant children from a high school in Guangdong Province (357 females; Mage = 15.3 at Time 1, SDage = 0.53) participated in this study. Random-intercept cross-lagged panel models (RI-CLPMs) were utilized to examine the longitudinal associations among family functioning, self-compassion, and emotion regulation in both groups. The results showed that, at the within-person level, family functioning reciprocally predicted self-compassion over time among migrant children, and it also exerted an indirect effect on emotion regulation, mediated by self-compassion. Among non-migrant children, emotion regulation positively predicted self-compassion over time, with no other observed cross-lagged effects.