Unlike normal huntingtin (htt) which is located predominantly in the cytoplasm, mutant htt is also found in the nucleus of affected neurons. Nuclear localization of toxic polyglutamine-containing proteins has been postulated to be necessary for the pathogenesis of triplet repeat disorders. However, little is known about the mechanism by which mutant htt enters the nucleus. We have recently reported exclusive nuclear localization of exon 1 mutant htt in striatal primary neuronal cultures from the HD94 conditional mouse model of HD. This seemed to contradict the predominant cytoplasmic localization of N-terminal htt reported from transfection experiments and prompted us to hypothesize that subcellular localization of the toxic htt fragment might be favoured in nondividing cells. To test this, we analyzed subcellular localization of mutant htt in HD94 mixed neuron-glia cultures. Subconfluent glial cells showed cytoplasmic localization. However, nuclear localization was prompted by confluence, by serum withdrawal, and by treatment with cell cycle progression inhibitors such as Ara C or lactacystin. BrdU labelling experiments further confirmed that nuclear localization does not occur in dividing cells. Our findings offer an explanation for the neuronal specific toxicity of mutant htt despite its ubiquitous expression. Unraveling the mechanism of this cell cycle arrest-dependent entrance into the nucleus may offer new opportunities for therapeutic intervention.