Adults with type 2 diabetes (T2D) often experience two common diabetes-related psychological distress: diabetes distress and depressive symptoms. Both are associated with adverse diabetes outcomes including poor self-management and glycemic control. However, diabetes distress and depressive symptoms differ in their associations with diabetes outcomes in T2D patients. This study proposes a hypothetical model to examine whether self-efficacy mediates the adverse effects of depressive symptoms and/or diabetes distress on self-care behaviors and glycemic control. Additionally, we examined the bi-directional relationships between diabetes distress and depressive symptoms to identify potential underlying mechanisms. This study conducted in 15 rural health clinics in Jiangsu province China. 900 adults with T2D participated in the prospective cohort study. The data Diabetes distress (the 17-item Diabetes Distress Scale, DDS17), depressive symptoms(the 10-item Center for Epidemiologic Studies Depression Scale, CESD-10), self-efficacy, self-care behaviors (diet and physical activity), metabolic variables (fasting plasma glucose, FPG) and demographic characteristics were assessed at baseline. Subsequent 12-month Hemoglobin A1C (HbA1c) were measured after baseline. Hierarchical multiple regression and bootstrap mediation analysis were used to test the effects and pathways among these associations. Of 843 participants (93.67%) of total cohort with available subsequent 12-month HbA1c levels, mean age was 66.08 years and 66.55% were women, 25.15% of them had depressive symptoms (CES-D ≥ 10), 12.20% had moderate diabetes distress (mean DDS ≥2) and 4.98% had the both two psychological distress. Hierarchical multiple regression showed higher DDS score significantly predicted unhealthy diet(β = -1.10, P < 0.001) but not physical activities, while CESD score was negatively associated with physical activity (β = -0.06, P < 0.001) but not diet. No independent effects of the two psychological distress variables on subsequent 12-month HbA1c were observed. Mediation analysis supported that elevated self-efficacy solely mediated the negative effect of both diabetes distress and depressive symptoms on diet (DDS score: β = -0.238, 95 BCE% CI [-0.350, -0.141]; CESD score: β = -0.010, 95 BCE% CI [-0.016, -0.005]), physical activities (DDS score: β = -0.446, 95 BCE% CI [-0.630, -0.283]; CESD score: β = -0.019, 95 BCE% CI [-0.030, -0.010]) and subsequent 12-month HbA1c (DDS score: β = 0.105, 95 BCE% CI [0.030,0.189]; CESD score: β = 0.004, 95 BCE% CI [0.001,0.009]). Additionally, the interplay of diabetes distress and depressive symptoms exerts their effects on diabetes outcomes directly and indirectly via self-efficacy. Self-efficacy may contribute to better diabetes outcomes and ameliorate negative effects of diabetes distress and depressive symptoms.