There is a need for standardisation in sampling and analysis of breath volatile organic compounds (VOCs) in order to minimise ubiquitous confounding effects. Physiological factors may mask concentration changes induced by pathophysiological effects. In humans, unconscious switching of oral and nasal breathing can occur during breath sampling, which may affect VOC patterns. Here, we investigated exhaled VOC concentrations in real-time while switching breathing routes. Breath from 15 healthy volunteers was analysed continuously by proton transfer reaction time-of-flight mass spectrometry during paced breathing (12 breaths min-1). Every two minutes breathing routes were switched (Setup-1: Oral → Nasal → Oral → Nasal; Setup-2: OralinNasalout → NasalinOralout → OralinNasalout → NasalinOralout). VOCs in inspiratory and alveolar air and respiratory and hemodynamic parameters were monitored quantitatively in parallel. Changing of the breathing routes and patterns immediately affected exhaled VOC concentrations. These changes were reproducible in both setups. In setup-1 cardiac output and acetone concentrations remained constant, while partial pressure of end-tidal CO2 (pET-CO2), isoprene and furan concentrations inversely mirrored tidal-volume and minute-ventilation. H2S (hydrogen-sulphide), C4H8S (allyl-methyl-sulphide), C3H8O (isopropanol) and C3H6O2 increased during oral exhalation. C4H10S increased during nasal exhalations. CH2O2 steadily decreased during the whole measurement. In setup-2 pET-CO2, C2H6S (dimethyl-sulphide), isopropanol, limonene and benzene concentrations decreased whereas, minute-ventilation, H2S and acetonitrile increased. Isoprene and furan remained unchanged. Breathing route and patterns induced VOC concentration changes depended on respiratory parameters, oral and nasal cavity exposure and physico-chemical characters of the compounds. Before using breath VOC concentrations as biomarkers it is essential that the breathing modality is defined and strictly monitored during sampling.