Despite the wide use of Caenorhabditis elegans (C. elegans) as a model organism, the first virus naturally infecting this organism was not discovered until seven years ago. The Orsay virus has a bipartite, positive sense RNA genome, with the RNA 1 segment encoding the RNA-dependent RNA polymerase (RdRP) and the RNA 2 segment encoding the capsid protein (CP) and another protein named delta (δ). δ can be expressed either as a free δ or a CP-δ fusion protein by ribosomal frameshift, but their structure and function were mostly unknown. To get a better understanding of the infection mechanism of this novel virus, I performed both structural and functional studies of the free δ protein as well as the CP-δ fusion protein. Using a combination of techniques including electron microscopy, X-ray crystallography, computational and biophysical analyses, we found that the Orsay δ protein forms a ~420-Å long, pentameric fiber with an N-terminal coiled coil, a β-stranded filament in the middle, and a C-terminal head domain. The N-terminal coiled coil of the polypeptide is essential for the proper folding and oligomerization of the pentameric structure. Interestingly, δ from the other two nematode viruses, namely Le Blanc and Santeuil viruses, also forms pentameric fibrous molecules similar to the Orsay δ. The relatively low amino acid sequence identities of ~30-40% among these δ proteins likely account for several notable structural differences among these fibers. For example, the internal globular domain observed in Orsay is not clearly visible in both Le Blanc and Santeuil fibers. Also, Santeuil fiber tends to bend in the middle, which does not apply to the other two δ fibers. On the other hand, we found that free δ is required for non-lytic egress of Orsay virus from worm intestinal cells. Transgenic worms over-expressing free δ showed that δ was preferably localized to the apical membrane of host intestine cells, where it likely facilitates virus release by reconstructing the terminal web network. The co-localization of δ and worm ACT-5 and their interaction suggest that Orsay δ may functionally interact with ACT-5 to mediate the viral exit from the worm intestine. For the CP-δ protein, both recombinant Orsay capsid containing CP-δ and purified Orsay virions are associated with protruding long fibers with globular heads at the distal end, suggesting that CP-δ protein can be incorporated into the Orsay capsid during assembly. To test the biological function of the capsid-associated CP-δ fibers, mutant viruses with disrupted fiber structures were generated by organism-based reverse genetics. These viruses were found to be either non-viable or poorly infectious according to phenotypic and qRT-PCR analyses. In addition, by performing a protein competition assay, we found that adding purified full-length δ protein into the culture medium could inhibit Orsay infection, while adding an N-terminal fragment δ(1-101) had no such effect. Based on the structure resemblance between the Orsay CP-δ fiber and the fibers from reovirus and adenovirus, we propose that CP-δ may interact with the host cellular receptors via its C-terminal head to mediate Orsay entry into worm intestine cells. Therefore, my study suggested that the Orsay δ protein likely has two independent functions at different stages of the virus life cycle. Considering the similarity between the worm and human intestinal cell morphologies, it is expected that my findings will lead to a better understanding of the infection mechanism of not only Orsay, but also gastrointestinal viruses infecting both mammals and humans.