Retraction of distal sensory axons is a prominent feature in diabetic peripheral neuropathy (DPN), a process amenable to insulin therapy. Nevertheless, diabetic patients and long-term diabetic mice develop motor deficits after longer durations of DPN, a process that may be related to insulin deficiency. To compare the efficacy of intranasal delivery of insulin (IN-I) and subcutaneous insulin (Subc-I) in preventing motor deficits in a long-term mouse model of DPN, IN-I or Subc-I, 0.87 IU daily or placebo was delivered in separate cohorts of diabetic and nondiabetic CD1 mice for 8 months. Radiolabeled detection was used to assess insulin delivery and biodistribution. Biweekly behavioral tests and monthly electrophysiological and multipoint quantitative studies assessed motor function deficits. Morphometric analysis of spinal cord, peripheral nerve, muscle innervation, and specific molecular markers were evaluated at and before the end point. Despite progressive distal axonal terminal loss, numbers and caliber of motor neurons were preserved. There were no differences in glycemia between IN-I and Subc-I-treated mice. Intranasal delivery of insulin and, to a lesser extent, Subc-I, protected against electrophysiological decline, loss of neuromuscular junctions, and loss of motor behavioral skills. Intranasal delivery of insulin was associated with greater preservation of the phosphatidylinositol 3-kinase signaling pathway involving Akt, cyclic AMP response element binding protein,and glycogen synthase kinase 3beta but did not alter extracellular signal-regulated kinase, mitogen-activated protein kinase/extracellular signal-regulated kinase, or c-Jun amino-terminal kinase. Thus, direct delivery of insulin to the nervous system might prevent motor deficit in human type 1 diabetes by preservation of the phosphatidylinositol 3-kinase-Akt pathway rather than only affecting glycemic levels; the effects of insulin on other signaling pathways may, however, play additional roles.