The detection of volatile organic compounds (VOCs) in human breath can be useful for the clinical routine diagnosis of several diseases in a non-invasive manner. Traditional methods of breath analysis have some major technical problems and limitations. Membrane extraction with a sorbent interface (MESI), however, has many advantages over current methods, including good selectivity and sensitivity, and is well suited for breath analysis. The aim of this project was to develop a simple and reproducible sampling device and method based on the MESI system for breath analysis. The feasibility and validity of the MESI system was tested with real human breath samples. Internal standard calibration methods were used for the quantitative analysis of various breath samples. Calibration curves for some main components (target analytes such as acetone and pentane) were determined in the research. The optimized stripping-side and feeding-side gas velocities were determined. The use of breath CO2 as an internal standard for the analysis of breath VOCs is an effective method to solve the difficulties associated with variations in the target analyte concentrations in a sample, which are attributed to mass losses and different breathing patterns of different subjects. In this study, the concentration of breath acetone was successfully expressed normalized to CO2 as in the alveolar air. Breath acetone of healthy males and females profiled at different times of the day was plotted using the MESI system, and results were consistent with the literature. This technique can be used for monitoring breath acetone concentrations of diabetic patients and for applications with other biomarker monitoring.