OBJECTIVE: We hypothesized that intranasal insulin (I-I) delivery targets the nervous system while avoiding potential adverse systemic effects when compared with subcutaneous insulin (S-I) for experimental streptozotocin-induced diabetic peripheral neuropathy (DPN). RESEARCH DESIGN AND METHODS: I-I or S-I at 0.87 IU daily or placebo were delivered in separate cohorts of diabetic and nondiabetic CD1 mice during 8 months of diabetes. Radiolabeled insulin detection was used to compare delivery and biodistribution for I-I and S-I. Biweekly behavioral testing and monthly electrophysiological and quantitative studies assessed progression of DPN. At and before end point, morphometric analysis of DRG, peripheral nerve, distal epidermal innervation, and specific molecular markers were evaluated. RESULTS: Radiolabeled I-I resulted in more rapid and concentrated delivery to the spinal cord and DRG with less systemic insulin exposure. When compared with S-I or intranasal placebo, I-I reduced overall mouse mortality and sensory loss while improving neuropathic pain and electrophysiological/morphological abnormalities in diabetic mice. I-I restored mRNA and protein levels of phosphoinositide 3-kinase/Akt, cyclic AMP response element-binding protein, and glycogen synthase kinase 3beta to near normal levels within diabetic DRGs. CONCLUSIONS: I-I slows the progression of experimental DPN in streptozotocin mice, avoids adverse effects associated with S-I treatment, and prolongs lifespan when compared with S-I. I-I may be a promising approach for the treatment of DPN.