Parkinson's disease appears to be a good candidate for gene therapy. The primary biochemical defect associated with the disease has been clearly determined as an absence of dopamine in the caudate-putamen, and the anatomical region where the neuropathologic hallmark of the disease, death of the nigral dopamine-producing neurons, occurs, remains circumscribed. Based on the biochemical and anatomical information gathered on Parkinson's disease, different gene therapy strategies have been devised to treat it. The first, and most explored strategy so far, consists in engineering cells to produce levodopa or dopamine so they will replace dopaminergic neurotransmission. Several types of cells have been employed in these experiments, and behavioral recovery has been reported in animal models of the disease. However, this approach cannot prevent neuronal death, nor reconstruct brain circuits. Another strategy is to protect cells by transferring genes that encode neurotrophic factors. Effort is now being concentrated into this research area, and promising results have recently been reported. Finally, an additional strategy aims at generating cells with a dopaminergic phenotype so they will be capable of replacing the missing dopaminergic neurons in biochemical, anatomical and functional terms. This has the potential to become an important constituent for an effective cure. Gene therapy holds significant promise for the treatment of neurodegenerative disorders, and Parkinson's disease treatment will benefit greatly from the knowledge and information arising from gene therapy research.