Non-small-cell lung cancer (NSCLC) accounts for 80% of lung cancer and has a 5 year survival rate of just 15%. Recently, tyrosine kinase inhibitors (TKIs) have been found to give dramatic therapeutic benefit to some patients increasing both their overall survival and quality of life compared with the more traditional and aggressive platinum-based chemotherapy. Subsequently, it was shown that EGFR and KRAS mutations act as biomarkers in predicting of patient response or lack thereof, respectively to TKIs. As such, the establishment of a clinically relevant assay for the detection of EGFR and KRAS mutations would contribute to personalised treatment of NSCLC patients to ultimately improve patient outcomes. However, there are many limitations and obstacles that limit the sensitivity of current protocols to detect these biomarkers, including the inherent heterogeneous nature of NSCLC tumours and the poor quality of genomic DNA due to biopsies being stored formalin-fixed, paraffin-embedded (FFPE) specimens. This study investigated the use of mineral oil to produce a high yield of quality DNA from FFPE NSCLC specimens. Additionally, significant optimisation experiments were carried out to find the optimum conditions for efficient amplification during PCR thermocycling using DNA derived from these FFPE specimens. Subsequently, the use of standard PCR, Co-amplification at lower denaturation temperature (COLD) PCR, restriction digest-mediated mutant enrichment PCR, single-stranded conformation polymorphism (SSCP) and DNA melt curve analysis were investigated as EGFR and KRAS mutation detection protocols in NSCLC FFPE specimens. Significantly, this study found that the use of mineral oil contributed to the extraction of a high yield of quality genomic DNA from the FFPE specimens and the use of a high fidelity DNA polymerase enzyme and a PCR buffer with a high magnesium concentration were required to produce amplifiable products from the FFPE specimens. Subsequently, it was found that both SSCP and DNA melt curve analysis could detect putative mutations in EGFR exon 21 and 19. Given that SSCP and DNA melt curve analysis could detect mutations in the NSCLC FFPE specimens, these mutation detection protocols were shown to be more sensitive than direct sequencing. However, further work is required to establish a clinical relevant mutation detection protocol for the routine detection of EGFR and KRAS mutations in FFPE NSCLC specimens.