A previous cross-sectional study found that problematic mobile phone use (PMPU) was associated with students' eating disorder symptoms. However, since the cross-sectional study cannot infer the causality and the direction of effect, the longitudinal relationship between the two and the mechanism behind this relationship are unclear. Therefore, the present study explores the prospective association between PMPU and eating disorder symptoms and related mediation mechanisms using a 1-year longitudinal study of 1,181 college students (from December 2019 [T1] to December 2020 [T2]). Survey tools used include the Mobile Phone Addiction Tendency Scale, the 10-item Connor-Davidson resilience scale, and the 12 item Short Form of the Eating Disorder Examination Questionnaire. The longitudinal relationship between PMPU and eating disorder symptoms and the mediating effect of resilience was analyzed using a cross-lagged model. The results showed that PMPU (β = 0.086, < 0.01) and resilience (β = -0.145, < 0.01) at T1 predicted eating disorder symptoms at T2, but not vice versa. PMPU was bidirectionally associated with resilience, and the prediction effect of PMPU at T1 to resilience at T2 (β = -0.151, < 0.001) was higher than the prediction effect of resilience at T1 to PMPU at T2 (β = -0.134, < 0.001). The standardized indirect effect of PMPU at T1 on eating disorder symptoms at T2 resilience was significant (β = 0.022, 95% = 0.010~0.040, < 0.001). Therefore, PMPU and resilience were predictive for eating disorder symptoms in college students, and resilience may play a mediating role in the prospective association between PMPU and eating disorder symptoms. This study provides new ideas and higher-level evidence for the development of prevention and intervention measures for college students' eating disorder symptoms.