Objective To analyze the developmental trajectories of middle school students'loneliness and social support,as well as to explore the interaction between loneliness and social support,so as to provide the evidence-based support for the mental health development of adolescents.Methods A total of 989 first-year students from four public middle schools in Xiangxi Tujia and Miao Autonomous Prefecture,Hunan Province were selected for three follow-up surveys by a cluster random sampling method(T1:March 2023,T2:June 2023,T3:December 2023).The UCLA Loneliness Scale-20(ULS-20)and Social Support Scale for University Students(SSSUS)were employed for questionnaire data collection.The growth mixture modeling was utilized to test the developmen-tal trajectories of loneliness and social support among middle school students,while the cross-lagged analysis was performed to in-vestigate their mutual influence.Results The scores for loneliness and social support in T1,T2 and T3 were(43.1±5.8,42.5±6.8,42.0±6.9;55.9±12.0,60.7±15.7,60.4±16.7),respectively.Correlational analysis revealed a significant negative correlation between loneliness levels(Tl,T2,T3)and social support(T1,T2,T3)(r=-0.47 to-0.36,P<0.01).Growth mixture modeling indicated a linear declining trend of middle school students'loneliness,and the developmental trajectory of social support showed a linear increasing trend,with significant individual differences in initial levels and rates of change(P<0.05).Cross-lagged analyses revealed that loneliness levels at T1 negatively predicted social support scores at T2(β=-0.16),and loneliness levels at T2 nega-tively predicted social support scores at T3(β=-0.12)(P<0.05).Additionally,prior loneliness positively predicted its subsequent levels,with path coefficients of 0.58 and 0.47,respectively(P<0.05).Social support scores at T1 negatively predicted loneliness levels at T2(β=-0.10),while scores at T2 negatively predicted loneliness levels at T3(β=-0.15)(P<0.05).Prior loneliness al-so positively predicted its subsequent levels,with path coefficients of 0.43 and 0.44,respectively(P<0.05).Conclusion The de-velopmental trajectory of middle school students'loneliness demonstrates a decreasing trend,while that of social support exhibits a linear increasing trend,indicating a longitudinal causal relationship between loneliness and social support.