In spite of the extensive studies performed on postmortem substantia nigra from Parkinson's disease patients, the aetiology of the disease has not yet been established. Nevertheless, these studies have demonstrated that, at the time of death, a cascade of events had been initiated that may contribute to the demise of the melanin-containing nigro-striatal dopamine neurons. These events include increased levels of iron and monoamine oxidase (MAO)-B activity, oxidative stress, inflammatory processes, glutamatergic excitotoxicity, nitric oxide synthesis, abnormal protein folding and aggregation, reduced expression of trophic factors, depletion of endogenous antioxidants such as reduced glutathione, and altered calcium homeostasis. To a large extent, the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and 6-hydroxydopamine (6-OHDA) animal models of Parkinson's disease confirm these findings. Furthermore, neuroprotection can be afforded in these models with iron chelators, radical scavenger antioxidants, MAO-B inhibitors, glutamate antagonists, nitric oxide synthase inhibitors, calcium channel antagonists and trophic factors. Despite the success obtained with animal models, clinical neuroprotection is much more difficult to accomplish. Although the negative studies obtained with the MAO-B inhibitor selegiline (deprenyl) and the antioxidant tocopherol (vitamin E) may have resulted from an inappropriate choice of drug (selegiline) or an inadequate dose (tocopherol), the niggling problem that still remains is why these drugs, and others, do work in animals while they fail in the clinic. One reason for this may be related to the fact that in normal human brains the number of dopaminergic neurons falls by around 3-5% every decade, while in Parkinson's disease this decline is greater. Brain autopsy studies have shown that by the time the disease is identified, some 70-75% of the dopamine-containing neurons have been lost. More sensitive reliable methods and clinical correlative markers are required to discern between confoundable symptomatic effects versus a possible neuroprotective action of drugs, namely, the ability to delay or forestall disease progression by protecting or rescuing the remaining dopamine neurons or even restoring those that have been lost.A number of other possibilities for the clinical failure of potential neuroprotectants also exist. First, the animal models of Parkinson's disease may not be totally reflective of the disease and, therefore, the chemical pathologies established in the animal models may not cause, or contribute to, the progression of the disease clinically. Second, because of the series of events occurring in neurodegeneration and our ignorance about which of these factors constitutes the primary event in the pathogenic process, a single drug may not be adequate to induce neuroprotection and, as a consequence, use of a cocktail of drugs may be more appropriate. The latter concept receives support from recent complementary DNA (cDNA) microarray gene expression studies, which show the existence of a gene cascade of events occurring in the nigrostriatal pathway of MPTP, 6-OHDA and methamphetamine animal models of Parkinson's disease. Even with the advent of powerful new tools such as genomics, proteomics, brain imaging, gene replacement therapy and knockout animal models, the desired end result of neuroprotection is still beyond our current capability.