Polyglutamine (polygln) expansion in specific proteins is one of the most intriguing pathogenic mechanisms causing adult-onset neurodegenerative disorders. In all the cases studied so far, the normal gene products tolerate a rather wide variation in size of a polygln tract (ranging typically between 10–35 glns) without any detectable adverse effect. However, beyond the threshold of about 35–42 glns those proteins acquire toxic properties, which correlate with a conformational change-revealed by the 1C2 antibody interaction—and a tendency of mutant proteins to aggregate in vitro and in vivo. To date nine polygln expansion disorders have been characterized, the list of which includes Huntington’s disease (HD) the most frequent of them, six spinocerebellar ataxias (SCA1-3, SCA6, SCA7, and SCA17), spino-bulbar muscular atrophy (SBMA) and dentatorubral-pallidoluysian atrophy (DRPLA) (reviewed in refs. 1 ,2 ). Most of theses diseases present a strong inverse correlation between the length of the polygln tract and the age of onset of clinical symptoms. At the DNA level, the expanded CAG repeats, which code for the polygln stretch, are found to be unstable upon transmission from one generation to the other, with a clear tendency to expansion. These two features account for the anticipation phenomenon (increased severity and earlier onset of the disease in successive generations) that is observed at various degrees in polygln expansion diseases. Because anticipation is observed in many other neurological disorders (and suggested in some non-neurological diseases), for which the causative gene has not yet been cloned, it is suggested that polygln expansion could be involved in the disease process (3 ,4 ).