Several dominant, late-onset neurodegenerative diseases (e.g. Huntington's disease) are caused by expansion of polyglutamine (polyQ) repeats within specific proteins. The diverse, yet overlapping, pathology of these diseases could be due to novel deleterious functions unique to each protein or to a common pathophysiology mediated by the long polyQ chains themselves. By engineering Drosophila to express different polyQ peptides, we find that expanded polyQ chains alone are intrinsically cytotoxic and cause neuronal degeneration and early adult death. We further find that this intrinsic toxicity is dependent on cell type and polyQ length and that the inclusion of other amino acids modifies and reduces toxicity. This is the first in vivo evidence that polyQs, when removed from their disease gene context, cause neurotoxicity. These studies provide a basis for understanding the diverse clinical presentations in terms of the intrinsic cytotoxic effect of polyQ peptides being modulated by protein context. Parallel experiments in which cytotoxic polyQ expansions were engineered into Dishevelled, a Drosophila protein containing a naturally occurring polyQ tract, strongly suggest that the effect of a toxic polyQ peptide can be neutralized by protein context. This animal model provides a simple and effective means of screening for therapeutics that relieves the polyQ-induced lethality, independent of any particular disease gene. By quantifying the degree of lethality in several transgenic lines, we have identified a number of genetically modified strains that are suitable for eventual testing of compounds or genes that ameliorate the pathology of polyQ peptides.