C-jun N-terminal kinase (JNK) regulates both the development of insulin resistance and inflammation. Podocytes of the widely used db/db mouse model of diabetic nephropathy lose their ability to respond to insulin as albuminuria develops, in comparison to control db/+ mice. Here we tested whether JNK inhibition or its gene deletion would prevent albuminuria in experimental diabetes. Phosphorylated/total JNK was significantly increased in vivo in glomeruli of db/db compared to db/+ mice. Treatment of podocytes isolated from these two strains of mice with tumor necrosis factor-alpha caused greater phosphorylation of JNK in those obtained from diabetic animals. When db/db mice were treated with a cell-permeable TAT-JNK inhibitor peptide, their insulin sensitivity and glycemia significantly improved compared to controls. We induced diabetes in JNK1 knockout mice with streptozotocin and found that they had significantly better insulin sensitivity compared to diabetic wild-type or JNK2 knockout mice. Albuminuria was, however, worse in all mice treated with the JNK inhibitor and in diabetic JNK2 knockout mice compared to controls. Nephrin expression was also reduced in JNK inhibitor-treated mice compared to controls. A similar degree of mesangial expansion was found in all diabetic mice. Our study shows that targeting JNK to improve systemic insulin sensitivity does not necessarily prevent diabetic nephropathy.