Volatile organic compounds (VOCs) released into the headspace air over human tissues infected with different bacteria were investigated in this work. The above-mentioned VOCs result both from bacterial metabolic processes (pathogen-specific signals) and from the matrix (tissue samples themselves). The objective of this study was to investigate whether one could reliably identify various microorganism strains that exist inside infected tissue samples by direct monitoring of the headspace atmosphere above their cultures. Headspace samples were directly interrogated using a GC-MS system, which produced distinct profiles for samples contaminated with single bacterial strains or with multiple strains (mixed infections). Principal component analysis (PCA) and predictive analysis based on receiver operating characteristics curves (ROC) were the statistical procedures utilized for differentiating between infected and uninfected samples, while network analysis and heat-mapping were used to highlight the connections between emitted volatiles and infectious pathogens. By using ROC curves, obtained results demonstrated that the area under the ROC (95% probability interval) was 0.86 in case of infected samples and 0.48 for uninfected samples. On the other hand, PCA highlighted separation between components coming from infected and uninfected patients, where 67% of variance was described from the first 2 principal components. The biomarker chemicals documented from this work, as well as the developed methodology may ultimately be applied to identify bacterial infections by analyzing exhaled breath.