A large body of evidence indicates that the progression of Parkinson's disease (PD) may be fast in the preclinical stage as well as during the first years of the disease, with a subsequent slowing down of the disease process. As has been shown in the Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism (DATATOP) study, the Unified Parkinson's Disease Rating Scale (UPDRS) motor examination scores declined at a rate of 8 to 9% per year in untreated patients. A subgroup of levodopa-naive DATATOP patients ("survivors") showed a much slower rate of progression, in the order of 3% per year, suggesting a more benign disease course. A number of clinical factors that may govern the rate of motor decline, such as age at onset, disease duration, gender, and clinical phenotype (akinetic-rigid versus tremulous) have been proposed; however, none of them is proven. In contrast, dopaminergic substitution undoubtedly has had a major impact on the natural history of PD, resulting in a reduction of the mortality ratio from about 3.0 to 1.5. This benefit has been noted particularly in patients in whom levodopa therapy was started early. The positive impact of levodopa is largely derived from its symptomatic action; its influence on the disease process itself remains controversial.