Epidemiological studies have suggested an etiologic relationship between pesticide exposure and Parkinson's disease (PD). Organochlorine pesticides were assayed in postmortem brain samples from 20 PD, 7 Alzheimer's disease (AD), and 14 nonneurological control cases. The three groups were similar in age at death, sex, and demographic variables. Only two of 16 pesticide residues screened were detected. A long-lasting residue of DDT (pp-DDE) was found in the majority of cases of PD and AD, as well as in all the control cases; pp-DDT was significantly more likely to be found in AD controls than the PD cases (Fisher's exact two-tailed, p = 0.04). Dieldrin was detected in 6 of 20 PD brains, 1 of 7 AD, and in none of 14 control samples. Despite the relatively small number of brains assayed, the association between Dieldrin and the diagnosis of PD was highly significant (p = 0.03). Dieldrin, a lipid-soluble, long-lasting mitochondrial poison, should be investigated as a potential etiological agent of Parkinsonism.