Glial cell line-derived neurotrophic factor (GDNF) and Neurturin (NRTN) bind to a receptor complex consisting of a member of the GDNF family receptor (GFR)-α and the Ret tyrosine kinase. Both factors were shown to protect nigro-striatal dopaminergic neurons and reduce motor symptoms when applied terminally in toxin-induced Parkinson's disease (PD) models. However, clinical trials based on intraputaminal GDNF protein administration or recombinant adeno-associated virus (rAAV)-mediated NRTN gene delivery have been disappointing. In this review, several factors that could have limited the clinical benefits are discussed. Retrograde transport of GDNF/NRTN to the dopaminergic neurons soma is thought to be necessary for NRTN/GFR-α/Ret signaling mediating the pro-survival effect. Therefore, the feasibility of treating advanced patients with neurotrophic factors is questioned by recent data showing that: (i) tyrosine hydroxylase-positive putaminal innervation has almost completely disappeared at 5 years post-diagnosis and (ii) in patients enrolled in the rAAV-NRTN trial more than 5 years post-diagnosis, NRTN was almost not transported to the substantia nigra pars compacta. In addition to its anti-apoptotic and neurotrophic properties, GDNF also interferes with dopamine homeostasis via time and dose-dependent effects such as: stimulation of dopamine neuron excitability, inhibition of dopamine transporter activity, tyrosine hydroxylase phosphorylation, and inhibition of tyrosine hydroxylase transcription. Depending on the delivery parameters, the net result of this intricate network of regulations could be either beneficial or deleterious. In conclusion, further unraveling of the mechanism of action of GDNF gene delivery in relevant animal models is still needed to optimize the clinical benefits of this new therapeutic approach. Recent developments in the design of regulated viral vectors will allow to finely adjust the GDNF dose and period of administration. Finally, new clinical studies in less advanced patients are warranted to evaluate the potential of AAV-mediated neurotrophic factors gene delivery in PD. These will be facilitated by the demonstration of the safety of rAAV administration into the human brain.