Three neurodegenerative diseases, Huntington's disease (HD), Kennedy's disease (hereditary spinobulbar muscular atrophy, SBMA), and type 1 spinocerebellar ataxia (SCA-1) have been found to share a common genetic defect: an unstable region of repeated CAG trinucleotides which are thought to be translated into a polyglutamine moiety. The unstable repeat regions occur near the N-termini of the predicted proteins for HD and SBMA, but the location of the CAG repeat region is not known for SCA-1. Each disease is notable for a relatively circumscribed region of central nervous system pathology, and the lack of predicted similarity of the abnormal proteins makes a common mechanism related to the function of each protein unlikely. In order to reconcile the similar genetic abnormalities with the disparities in phenotypes, we suggest a common thread with regard to the pathogenesis of neuronal death. We hypothesize that the mechanism of neurotoxicity in these diseases occurs not through the production of abnormal proteins, but by the generation of abnormal posttranslational cleavage products. These products, in part consisting of abnormally large polyglutamine moieties, act to disturb the cellular and mitochondrial milieu such that energy metabolism is impaired, rendering specific regions of the nervous system vulnerable, and resulting in the clinical phenotypes of HD, SBMA, and SCA-1. We offer this interpretation of recent genetic findings from a neurobiologic perspective, in addition to suggesting testable hypotheses concerning potential disease mechanisms.