There are currently no disease modifying treatments available for frontotemporal dementia (FTD). Pathological heterogeneity within and between FTD phenotypes and genotypes makes accurate diagnosis challenging. Biomarkers that can aid diagnosis and monitor disease progression will be critical for clinical trials of potential treatments. Positron emission tomography (PET) imaging provides insights into molecular changes in the brain during life that are otherwise only directly quantifiable at postmortem. In this thesis I aimed to identify potential biomarkers of FTD using PET imaging. In Chapter 3 I use PET imaging of glucose metabolism to identify early neuronal dysfunction in presymptomatic genetic FTD, revealing specific involvement of the anterior cingulate in a subgroup of mutation carriers. In Chapter 4 I evaluate the utility of a PET tracer of tau protein deposition in genetic FTD against volumetric imaging, which appears to provide a more sensitive biomarker of disease than this tau PET tracer in FTD. In Chapter 5 I investigate neuroinflammation via PET imaging and identify different areas of neuroinflammation in different FTD genotypes, suggesting an association between neuroinflammation and protein deposition and that PET imaging of neuroinflammation might provide a sensitive biomarker in MAPT-related FTD. In Chapter 6 I investigate synaptic and mitochondrial dysfunction via PET imaging in FTD, the latter of which has been previously unexplored. I reveal marked differences in both markers in FTD versus controls which suggests both might provide sensitive biomarkers of disease. Furthermore, in Chapter 7 I evaluate the same biomarkers at longitudinal follow up where I find continued reductions in mitochondrial function over time suggesting mitochondrial PET imaging may provide a biomarker of disease progression in FTD. Future replication of the findings in this thesis in larger cohorts might facilitate the advancement of clinical trials in FTD.