Pulmonary infections remain the leading cause of communicable deaths worldwide. Yet, despite their prevalence the diagnosis of bacterial lower respiratory tract infections remains a challenge reliant on non-specific symptomology and culture-based methodologies that have limitations. As such, the early, accurate and reliable diagnosis of bacterial infections is troublesome, often leaving clinicians with a difficult decision to ‘treat or not to treat’ with antibiotics. Efforts to modernise clinical microbiology and improve pathogen detection are needed and reflects the subject of this thesis in which two alternative approaches were explored. The first part of the thesis focused on the clinical application of 16s rRNA gene sequencing in patients with cystic fibrosis or suspected ventilator associated pneumonia. Both direct pathogen detection and ecological markers of bacterial composition determined using 16s sequencing were evaluated against conventional cultures to gain further insight into the clinical role the microbiome may have. The second part describes the development of a preclinical in vitro model aimed to discover potential diagnostic volatile organic compounds (VOCs). The model was used to evaluate the individual contributions of oxidative stress and bacterial metabolism on VOC production before being translated to exhaled breath studies in the hope to validate potential biomarkers for pulmonary infections. Whilst both methods demonstrated a potential application, their clinical integration into everyday practice is challenged by limitations outlined in this thesis that must be addressed before their use can be recommended.