The membrane interphase probe (MIP™) from Geoprobe Systems® has frequently been applied in different countries for the characterization of soil contaminated with volatile organic carbons (VOCs). Experience shows that misinterpretation of the collected data is common. This is mainly due to the lack of understanding and knowledge related to the detectors used, their detection limits, and the sensitivity of the MIP system. It has been noticed that the sensitivity of the system given by the producer and by different users are rather optimistic, e.g., the values given are lower (= better) than those actually experienced in the field. A need for a better understanding of the MIP system sensitivity, combined with a more scientifically based interpretation of the collected data, exists. Both laboratory tests (using solutions) as well as field measurements were carried out using different detector configurations to allow a better interpretation of the detector signals/system sensitivity and to collect qualitative information. These configurations were: (1) detectors stand alone; (2) the use of a 2-ml sample loop, and (3) a purge and trap system. The configurations (2) and (3) are used in combination with a capillary column to carry out on-site qualitative and semiquantitative analyses. With respect to the configuration of “detectors stand alone,” detection limits for toluene (in aqueous solutions) range between 4 ppm (flame ionization detector—FID) and 10 ppm [photo ionization detector (PID)]. For chlorinated aliphatic hydrocarbons (CAHs), observed limits are 10 ppm (FID), 4–50 ppm (PID), and 3–10 ppm [dry electrolytic conductivity detector (DELCD)]. When using the 2-ml sample loop, relatively high concentrations have to be initially present in the soil. Observed detection limits for mono-aromatic hydrocarbons are 5–100 ppm; for CAHs, 1–50 ppm; for alkanes, 250–400 ppm; and for MTBE, 25 ppm. The application of purge and trap results in a better resolution and the detection of lower concentrations. Consequently, a better identification of the pollution with depth is possible. In this case, the detection limits are a function of the concentrations and the flushing time. In relation to the qualitative analyses, it was found that the configuration of the MIP-system with the built-in capillary column and the 2-ml sample loop or the purge and trap preconcentrator, respectively, are useful to carry out on-site analyses, thus allowing a better identification of the pollution in a vertical profile. The measurements carried out using the MIP with detectors stand-alone or in combination with a loop or trap, or connected to a column, confirm that analysis is indeed very useful to characterize VOC source zones when knowing and understanding its performance. This relates mainly to the detection limits of the MIP system. For a selection of parameters, such limits have been obtained. These values seem to be more realistic than those found in the few references where numbers are given. For the qualitative measurements, it can be concluded that a better resolution is obtained, and pollutants present in lower concentrations will be detected when using the purge and trap. It is advised to determine the optimal flushing time and the detection limit of the expected pollutants in advance. This study indicates that there is still a need for further measurements and discussion between users. Finally, additional data should result in a better interpretation of the collected field data.