While many pathophysiologic pathways may be involved in the neurodegeneration responsible for PD, genetic factors are likely to determine a general susceptibility to neurodegeneration. There are a number of genetic polymorphisms of genes such as those coding for the cytochrome p450 super-family of genes referred to as 'susceptibility genes'. However, they are generally not sufficient to cause disease unless a person encounters exposure to an environmental toxin: the disease is caused by a gene-environment interaction. Thus, it is imperative to assess genetic susceptibility in individuals exposed to a toxin. We will test the gene-environment interaction hypothesis by conducting an epidemiologic population-based case-control study of newly diagnosed PD patients from three rural California counties: Kern, Fresno, and Tulare. Over a four year period, we expect to collect 400 cases referred to us by local neurologists, farm worker clinics, and Parkinson's foundations. For each case, one population control will be selected at random from residential parcel maps and Medicare databases and, in addition, one unaffected sibling control and - when possible - affected siblings to avoid potential biases and inefficiencies inherent in the use of each type of control. For each study subject, an environmental and occupational pesticide exposure estimate will be derived using California pesticide-use reporting (PUR) data and information about pesticide application on crops in combination with crop patterns shown in satellite images and aerial photographs; in addition, extensive exposure interviews will be conducted with all study subjects. In a three-tiered approach to examine the effects of gene-environment interactions we will: 1) test for association (and linkage) of PD to selected loci associated with PD in earlier studies using multiallelic repeat markers and genotyping; 2) test for association using intragenic single nucleotide polymorphisms (SNPs) of 50 candidate genes arrayed to create "the PD array"; and 3) use future technical possibilities to screen for genome wide associations using array technology to scan 5,000-10,000 SNPs throughout the genome. Data analysis will employ hierarchical modeling procedures to take into account multiple comparison issues and to incorporate prior knowledge such as increased neurotoxicity due to the interaction of gene products and chemicals.