Clay has been used for many hundreds of years in building dikes. Their physical properties and thickness are essential for a dike to maintain its function. For the last fifteen years electromagnetic induction instruments (EMI) have been playing a growing role in mapping electromagnetic soil properties. Layered models with sharp boundaries between different soil units represents the soil profile accurately. Therefore, two database inversion procedures are developed and proposed. Synthetic data sets and field measurements have been used to validate and compare our retrieved electrical conductivity/resistivity models. Field measurements of a so-called multi-receiver EMI instruement (CMD-Explorer and CMD-MiniExplorer) have been used to perform the database inversion on, and are being compared to ERT data and hand drill data. The EMI instruments (CMD-Explorer and CMD-MiniExplorer) consisting of three receiving antennas, fixed operating frequency and two measurement configurations will not allow us more information than of a estimate three-layered model. Therefore, the calibration process here proposed is based on finding the best fit of the electromagnetic responses (in the least squares sense) in a 5-D solution space created by calculating the electromagnetic responses using five medium parameters and the layered earth response. These medium parameters are the three layered conductivity’s and thicknesses of the first two layers. Using these five medium parameters, a database is built based on predefined ranges of medium parameters and is being used to perform the inversion procedure with. Two methods of inversion have been proposed: constrained and full database inversion. The synthetic models have been recovered with an average data misfit of 0.0096 and 0.0032 for the constrained data base inversion for respectively model 1 and model 2. For the full database inversion data misfit values of 0.00316 and 0.0021 were accomplished. Comparing ERT and electrical resistivity obtained from EMI measurements, one could see that the trend and first two layer thicknesses were recovered. Verifying EMI measurements using hand drill data, we were able to recover the top layer at few locations.