Sensory polyneuropathy can be a serious problem, but for the majority of clinically important neuropathies there are no available therapies. Neurotrophic and neuroprotective peptide factors have been identified that prevent or reverse neuropathy in rodent models of disease, but delivery of these highly pleiotropic peptides has posed an obstacle for translation into effective human therapies. Gene transfer into muscle using viral or non-viral vectors, or into neurons of the dorsal root ganglion using herpes simplex virus-based vectors, provides an alternative means to achieve this end. Studies in animal models have been promising, and the first human trial, using a plasmid to transfer the gene coding for vascular endothelial growth factor into muscle for the treatment of diabetic neuropathy, is now underway. Evidence supporting the trial and the challenges facing this therapy are reviewed.