DNA polymerases play a fundamental role in the transmission and maintenance of genetic information and have become an important in vitro diagnostic and analytical tool. The Loop-mediated isothermal DNA amplification (LAMP) method has major applications for disease and pathogen detection and utilises the unique strand-displacement activity of a small group of thermostable DNA polymerases. The Large (Klenow-like) Fragment of Geobacillus stearothermophilus DNA polymerase I (B.st LF Pol I) currently serves as the enzyme of choice for the majority of these isothermal reactions, with few alternatives commercially available. An increasing need for point-of-care nucleic acid diagnostics is now shifting detection methods away from traditional laboratory based chemistries, such as the polymerase chain reaction (PCR), in favour of faster, and often simpler, isothermal methods. It was recognised that in order to facilitate these rapid isothermal reactions there was a requirement for alternative thermostable, strand-displacing DNA polymerases and this was the basis of this thesis. This thesis reports the successful identification of polymerases from Family A, chosen for their inherent strand-displacement activity, which is essential for the removal of RNA primers of Okazaki fragments during lagging-strand DNA synthesis in vivo. Twelve thermophilic organisms, with growth temperature ranges between 50oC and 80oC, were identified and the genomic DNA extracted. Where DNA sequences were unavailable, a gene-walking technique revealed the polA sequences, enabling the Large Fragment Pol I to be cloned and the recombinant protein over-expressed in Escherichia coli. A three-stage column chromatography purification permitted the characterisation of ten newly identified Pol I enzymes suitable for use in LAMP. Thermodesulfatator indicus (T.in) Pol I proved to be the most interesting enzyme isolated. Demonstrating strong strand-displacement activity and thermostability to 98oC, T.in Pol I is uniquely suitable to a newly termed heat-denaturing LAMP (HD-LAMP) reaction offering many potential advantages over the existing LAMP protocol. The current understanding of strand-displacement activity of Pol I is poorly understood. This thesis recognised the need to identify the exact regions and motifs responsible for this activity of the enzyme, enabling potential enhancements to be made. Enzyme engineering using site-directed mutagenesis and the formation of chimeras confirmed the importance of specific subdomains in strand-separation activity. With this knowledge, a unique Thermus aquaticus (T.aq) Pol I mutant demonstrated sufficient strand-displacement activity to permit its use in LAMP for the first time. The fusion of Cren7, a double-stranded DNA binding protein, to Pol I for use in LAMP is also reported. Although the fusion construct was found to reduce amplification speed, enhancements were observed in the presence of increased salt concentrations and it is suggested here as a means for future enzyme development.