This research was initiated by an outbreak of fowl adenovirus (FAdV) associated inclusion body hepatitis (IBH) in South Africa (SA) during 2008. The fowl adenoviruses involved in this outbreak could be identified by restriction enzyme fragment length polymorphisms (RFLP) and sequencing of the PCR amplification products from the FAdV L1 hexon loop. The relationship of these strains to the International Committee on Taxonomy of Viruses (ICTV) reference strains for FAdV could be established by phylogenetic analysis. The SA FAdV isolates showed close relationship (99 %) to the ICTV reference strain T8-A and 764 for FAdV-8b and the reference strain P7-A for FAdV-2. Although a complete epidemiological study was not performed data obtained from the phylogenetic analysis data also suggested that the fowl adenoviruses involved in this outbreak of IBH might have been introduced into the country. A dose of 106.00 EID50 /mℓ virulent FAdV-2 and 105.97 EID50 /mℓ virulent FAdV-8b was sufficient to cause 80-87 % mortality rates for embryos challenged with FAdV-2 and 65-80 % mortality in SPF embryos challenged with FAdV-8b. v Fowl adenovirus type-specific antibodies are masked by group-specific antibodies in ELISA. The L1 hexon loop of the fowl adenoviral capsid contains type-specific epitopes located between group-specific regions which could be used for development of a type-specific ELISA. A novel approach to include additional type-specific FAdV epitopes to select for type-specific antibody binding was followed during the development of the ELISA described in this study. A dimeric protein which targets the type-specific region within the L1 loop region of the FAdV-2 and FAdV-8b hexon was designed to include additional type-specific epitopes. Amino acid alignment of this region showed less than 46 % homology which presented an opportunity to investigate its use as coating antigen to detect type-specific antibodies in an indirect ELISA. The purified expression products of dimeric codon optimised genes encoding the variable regions within the FAdV L1 hexon loop for FAdV-2 and FAdV-8b were used as coating antigen in ELISA. The assay conditions were optimised with the Taguchi method for optimisation of experiments with multiple variables. The diagnostic performances of the ELISA were evaluated using 100 serum samples from vaccinated birds and birds with no previous history of exposure. The assay was able to detect type-specific antibodies with an overall assay accuracy of 85.1 % for FAdV-2 and 92.3 % for FAdV-8b. An embryo challenge model to measure the ability of maternal antibodies to protect against challenge with virulent FAdV was developed in this study. This challenge model was supported by macroscopical, histopathological and PCR data and was sensitive enough to be used for vaccine efficacy studies. A comparative study to evaluate the performance of formalin inactivated autogenous vaccine which contained whole virus to a fiber subunit vaccine which contained insoluble and refolded fiber proteins of both FAdV-2 and FAdV-8b. Synthetic genes encoding the complete fiber proteins of both FAdV-2 and FAdV-8b were cloned and expressed in E. coli. Both the fiber proteins were insoluble but were used as crude extracts in an experimental vaccine for vaccination of SPF birds. Purified refolded fiber protein fractions were also prepared from these insoluble fractions and were used for vaccination of another group of SPF parent birds. The autogenous bivalent formalin inactivated FAdV vaccine completely protected embryos from vaccinated parent birds against in ovo challenge with both FAdV-2 and FAdV-8b. Whilst the insoluble FAdV-8b fiber protein subunit vaccine protected against challenge, the FAdV-2 fiber protein did not. Vaccine prepared from purified refolded fiber proteins of FAdV-2 and FAdV-8b did not protect embryos from vaccinated parents upon in ovo challenge. / Thesis (PhD)--University of Pretoria, 2013. / gm2014 / Veterinary Tropical Diseases / Unrestricted