The advanced non-invasive diagnostic method is an analysis of exhaled volatile organic compounds (VOCs), which are regarded as biomarkers of various diseases. A gravimetric-type electronic nose (e-nose) has shown great promise in exhaled VOC-pattern identification, and its efficiency is determined by the adsorption capacity of the sensing materials towards a specific VOC. The present study employed molecular dynamics (MD) calculations with the OPLS all-atom force field to investigate the adsorption of an isolated molecule of volatile fatty acids, acetone, ethanol, methanol, acetic aldehyde, neopentane, and ethane in a unit cell of the lanthanum-based metal-organic framework (La-BTC MOF), which was proposed as a potential sensing material for the gravimetric-type e-nose. The adsorption potential determined as a change in the potential energy, ΔEp, of the total simulation system upon the penetration of a single molecule into a unit pore of La-BTC MOF was calculated for all the targeted VOCs and water to be 1.56–10.78 kJ·mol−1, following the order: ethane < neopentane < acetic aldehyde < ethanol < methanol < water < acetone < butanoic acid < caproic acid. Therefore, the biomarkers with ΔEp > ΔEp(water) ∼ 7.3 kJ/mol can be captured from the exhaled breath. According to the MD calculations, an oxygen-containing biomarker molecule in the adsorbed state was oriented by its oxygen atom towards the La3+ cation, so that the O---La3+ distance of 0.28–0.32 nm was the shortest among those for other constituent atoms. Thus, the strong electrostatic interactions between the oxygen atoms of the adsorbate molecules and the La3+ cation of the adsorbent were most responsible for the capture of the oxygen-containing biomarkers by the La-BTC MOF. The competing adsorption effect between water and oxygen-containing biomarker molecules was examined in dependence on the water content and the way the water molecules were introduced into the simulation system.