For easy handling and speed of lung diseases diagnostics, approaches based on volatile organic compounds (VOCs), including those emitted by pathogenic microorganisms, are considered but currently require considerable sampling efforts. We tested whether easy-to-handle and fast detection of lung infections is possible using solid-phase microextraction (SPME) of 100 ml of exhaled breath. An analytical procedure for the detection of VOCs from the headspace of epithelial lung cells infected with four human pathogens was developed. The feasibility of this method was tested in a cystic fibrosis (CF) outpatient clinic in vivo. Exhaled breath was extracted by SPME and analyzed by gas chromatography-mass spectrometry (GC-MS). The compositions of VOCs released in the infection model were characteristic for all individual pathogens tested. Exhaled breath of CF patients allowed clear distinction of CF patients and controls by their VOC compositions using multivariate analyses. Interestingly, the major specific VOCs detected in the exhaled breath of infected CF patients in vivo differed from those monitored during bacterial in vitro growth. SPME extraction of VOCs from 100 ml of human breath allowed the distinction between CF patients and healthy probands. Our results highlight the importance of assessing the entire pattern of VOCs instead of selected biomarkers for diagnostic purposes, as well as the need to use clinical samples to identify reliable biomarkers. This study provides the proof-of-concept for the approach using the composition of exhaled VOCs in human breath for the rapid identification of infectious agents in patients with lower respiratory tract infections.