Machado-Joseph disease (MJD) is the most common dominant spinocerebellar ataxia. MJD is caused by a CAG trinucleotide expansion in the ATXN3 gene, which encodes a protein named ataxin-3. Ataxin-3 has been proposed to act as a deubiquitinating enzyme in the ubiquitin-proteasome pathway and to be involved in transcriptional repression; nevertheless, its precise biological function(s) remains unknown. To gain further insight into the function of ataxin-3, we have identified the Caenorhabditis elegans orthologue of the ATXN3 gene and characterized its pattern of expression, developmental regulation, and subcellular localization. We demonstrate that, analogous to its human orthologue, C. elegans ataxin-3 has deubiquitinating activity in vitro against polyubiquitin chains with four or more ubiquitins, the minimum ubiquitin length for proteasomal targeting. To further evaluate C. elegans ataxin-3, we characterized the first known knockout animal models both phenotypically and biochemically, and found that the two C. elegans strains were viable and displayed no gross phenotype. To identify a molecular phenotype, we performed a large-scale microarray analysis of gene expression in both knockout strains. The data revealed a significant deregulation of core sets of genes involved in the ubiquitin-proteasome pathway, structure/motility, and signal transduction. This gene identification provides important clues that can help elucidate the specific biological role of ataxin-3 and unveil some of the physiological effects caused by its absence or diminished function.