Presymptomatic detection of Parkinson's disease is necessary if neuroprotective therapies are to be utilized in its treatment. Various methods (PET, electrophysiology, enzyme assays, olfactory function) may be applicable but none has been rigorously evaluated. Other possible approaches are now considered. Plasma HVA levels (pHVA) in the presence of debrisoquine may reflect cerebral dopamine function. However, there are no detectable differences in pHVA between newly diagnosed and untreated parkinsonian patients and control subjects. Compensatory increases in dopamine turnover may mask a decrease in pHVA in the early stages of the disease. So, at present this technique could not be used as a diagnostic tool. Post-mortem studies of brain in Parkinson's disease may provide clues to biochemical markers indicative of nigral pathology. Mitochondrial complex I activity is reduced in substantia nigra in Parkinson's disease and it was reported also to be markedly reduced in blood platelets. However, subsequent studies suggest that the difference in platelet complex I activity is too small to be diagnostic of Parkinson's disease. There are also selective reductions in brain glutathione levels in Parkinson's disease restricted to substantia nigra, which do not occur in other neurodegenerative disorders and are not due to drug treatment. Importantly, in incidental Lewy body disease (preclinical Parkinson's disease) nigral glutathione levels are reduced to the same degree as in advanced Parkinson's disease. So, some peripheral index of altered glutathione function may be valuable in the early detection of the disease process.