Background Depression is associated with poor insulin sensitivity. We evaluated the long‐term effects of a cognitive behavioral therapy (CBT) program for prevention of depression on insulin sensitivity in adolescents at risk for type 2 diabetes (T2D) with depressive symptoms. Methods One‐hundred nineteen adolescent females with overweight/obesity, T2D family history, and mild‐to‐moderate depressive symptoms were randomized to a 6‐week CBT group (n = 61) or 6‐week health education (HE) control group (n = 58). At baseline, posttreatment, and 1 year, depressive symptoms were assessed, and whole body insulin sensitivity (WBISI) was estimated from oral glucose tolerance tests. Dual energy X‐ray absorptiometry assessed fat mass at baseline and 1 year. Primary outcomes were 1‐year changes in depression and insulin sensitivity, adjusting for adiposity and other relevant covariates. Secondary outcomes were fasting and 2‐hr insulin and glucose. We also evaluated the moderating effect of baseline depressive symptom severity. Results Depressive symptoms decreased in both groups (P < .001). Insulin sensitivity was stable in CBT and HE (ΔWBISI: .1 vs. .3) and did not differ between groups (P = .63). However, among girls with greater (moderate) baseline depressive symptoms (N = 78), those in CBT developed lower 2‐hr insulin than those in HE (Δ‐16 vs. 16 μIU/mL, P < .05). Additional metabolic benefits of CBT were seen for this subgroup in post hoc analyses of posttreatment to 1‐year change. Conclusions Adolescent females at risk for T2D decreased depressive symptoms and stabilized insulin sensitivity 1 year following brief CBT or HE. Further studies are required to determine if adolescents with moderate depression show metabolic benefits after CBT.