The RNA interference (RNAi) repertoire of metazoans is principally composed of three independent but related systems -endogenous small interfering RNAs (endosiRNAs), micro RNAs (miRNAs) and Piwi-interacting RNAs (piRNAs). Although the endosiRNA pathway has been demonstrated in all five eukaryotic supergroups, the miRNA and piRNA pathways of metazoans likely evolved subsequent to metazoan divergence from other eukaryotes. These innovations evolved in a 100-200 million year period between the emergence of animal multicellularity and the divergence of the sponges. It is not yet known if the organisation and function of these systems, as characterised in bilaterian model species (vertebrates, Drosophila and C. elegans), is shared between early branching phyla (Porifera, Ctenophora, Cnidaria and Placozoa) and bilaterians. Preliminary evidence is mixed with commonalities like the existence of 'ping-pong' piRNA biogenesis in sponges and cnidarians contrasted with the lack of sequence and secondary structure homology of miRNAs between sponges and bilaterians.This thesis focuses on the RNAi systems of three early branching metazoan phyla that represent critical branches in early animal evolution (ie. Porifera, Ctenophora and Cnidaria). As no such tool was available, I developed a bioinformatic pipeline to annotate the RNAi components of mapped small RNA libraries from the demosponge Amphimedon queenslandica, the ctenophore Mnemiopsis leidyi and the cnidarian Nematostella vectensis. Detailed in Chapter 2, this pipeline, named RNAiTool, clusters mapped small RNAs from high-throughput sequencing data and then annotates those clusters based on a simple set of criteria, primarily focused on the read-length of the constituent small RNAs.RNAiTool was also used in the annotation of small RNA datasets from the well-studied bilaterian Drosophila melanogaster, providing a point of comparison for the less wellunderstood non-bilaterian species. As well as interrogating individual datasets, RNAiTool can be used to compare the expression dynamics of endo-siRNA and piRNA cluster between datasets. Although it was used here to investigate the expression dynamics of endo-siRNA and piRNA cluster through development, RNAiTool can be used to assess the expression dynamics of clusters between any test and control datasets.The second major innovation presented in this thesis is the introduction of the uniformity index. This property of small RNA clusters describes the uniformity of small RNA coverage across the length of that cluster. This simple index provides a rapid method 3 for the identification of potential miRNAs and other highly expressed small RNA producing loci from large pools of small RNA clusters. This thesis takes the first steps towards an understanding of the evolution and function of RNAi systems in early branching metazoans, in particular the demosponge A.queenslandica. It is hoped that the framework for RNAi annotation developed here will prove useful for other studies of emerging model RNAi systems. 4