Pediatric obesity is a significant clinical and public health issue for African American girls in which low physical activity (PA) is a contributor. The mother-daughter relationship (MDR) has rarely been examined in the context of improving health behaviors such as PA and mental health outcomes (MHO) within this population. PURPOSE: To examine if change in PA following a 12-week culturally-tailored mother-daughter PA intervention predicts change in MHO variables (self-esteem, depressive symptoms, body image dissatisfaction) and MDR in pre-adolescent African American girls. METHODS: Mothers (n=27; age=36.0±17.0 years; body mass index (BMI)=34.0±7.4 kg/m2) and daughters (n=27; age=9.0±1.4 years; BMI=20.3±5.7 kg/m2, BMI percentile=73%) randomized to the mother-daughter dance group were examined in this analysis. Physical activity levels were assessed with Actigraph GT3X+ accelerometers for seven days and validated questionnaires. Mental health outcome variables and MDR were assessed using validated questionnaires. Spearman correlations were used to examine associations between variables. MANOVA was used to assess differences in PA levels across three time points. Paired t-tests and ANOVA were used for MHO variables and MDR across two and three time points, respectively. Simple regression was used to assess if PA self-efficacy and MDR mediated changes in PA. Hierarchical linear modeling (HLM) assessed if changes in PA variables predicted changes in MHO variables. RESULTS: Several significant correlations were observed at baseline and post-intervention such as the negative relationships between daughters’ light PA (% time) and depressive symptoms as well as a daughters’ BMI percentile and body image dissatisfaction. Significant reduction was observed in daughters’ self-reported PA (p=0.04) pre- to post-intervention. No other significant changes were observed. Change in PA did not predict change in MHO variables, but there was a negative effect of average BMI percentile on self-esteem (p=0.017) and body image dissatisfaction (p=0.002). CONCLUSION: In this sample of pre-adolescent African American girls, change in objectively measured PA did not predict change in MHO. The lack of significant findings could be attributed to low attendance of the intervention. Future studies should examine these relationships in a larger sample and explore the use of technology to combat low attendance.