RNAi is widely used as a genetic tool in a range of model organisms, however still relatively little is known about the endogenous functions of RNAi. In this thesis we have studied several aspects of the endogenous role of RNAi in order to enhance our understanding of the capabilities of this mechanism. We investigated an RNAi-related silencing mechanism (cosuppression) which is triggered by repetitive sequences resulting in silencing of endogenous genes in trans. We found both chromatin remodeling factors and RNAi factors to be involved in cosuppression, indicating a transcriptional branch to the RNAi mechanism. We also identified three members of a relatively new family of RNA-metabolising enzymes in C. elegans: the TRF-like nucleotidyltransferases. Analysing the family of TRF-like nucleotidyltransferases we found that all members are expressed in the germline of C. elegans and that three of them (cde-1, cde-2 and rde-3) are required in different combinations of RNAi-related silencing processes. Nucleotidyltransferase mutants cde-1 and rde-3 showed defects in germline stability and were unable to maintain germline function over time, suggesting that the RNAi silencing pathway and developmental pathways of the germline intersect at the level of the TRF-like nucleotidyltransferases. Next we created tagged constructs of RDE-3 to further analyse its function, and found that RDE-3 is localized to the germline cytoplasm, and to a subset of neurons. The protein was detected in two complexes: a large complex of 600-700kD, and a smaller less abundant 400kD complex. We used mass spectrometry to identify potential binding partners of RDE-3. We also studied the function of CDE-1 in RNAi and chromosome segregation. CDE-1 localises specifically to mitotic chromosomes in embryos, and requires the RdRP EGO-1, and the Argonaute protein CSR-1 for this localisation. We found that CDE-1 is required for the uridylation of CSR-1 bound siRNAs, and that in the absence of CDE-1 these siRNAs accumulate to inappropriate levels, accompanied by defects in both meiotic and mitotic chromosome segregation. Elevated siRNA levels are associated with erroneous gene silencing, most likely through the inappropriate loading of CSR-1 siRNAs into other Argonaute proteins. We propose a model in which CDE-1 restricts specific EGO-1-generated siRNAs to the CSR-1 mediated, chromosome associated RNAi pathway, thus separating it from other endogenous RNAi pathways. The conserved nature of CDE-1 suggests that similar sorting mechanisms may operate in other animals, including mammals. Finally, we identified a homologue of the small RNA methylating enzyme HEN-1, named REM-1. We showed that rem-1 mRNA is present in the germline, and that REM-1 is responsible for the methylation of the class of 21U RNAs confirming the notion that 21U RNAs are the worm counterpart of piRNAs. Methylation of 21U RNAs, however seems to be not essential for overall 21U RNA function. Taken together we identified two different chromatin-related branches to the RNAi mechanism, which changes our view of RNAi from a posttranscriptional silencing mechanism to a mechanism with a wider range of effects. In addition we reveal a regulatory mechanism involving nucleotidyltransferases and methylases, determining the stability and Argonaute loading preferences of small RNA.