Dielectric permittivity is a fundamental parameter for all investigations based on electromagnetic waves, particularly for moisture sounding. However, the physical principles of dielectric permittivity in natural materials and its influence on propagation velocity, the reflection coefficient, and the decay of the electromagnetic waves have been rarely addressed for variable substrate compositions or frequencies. Hence, this paper aims to fill this gap by systematically investigating quantitative relationships between dielectric permittivity and soil water content for a continuum of the three most common soil components in moderately temperate humid latitudes (i.e., illite, carbonate, and quartz‐rich sand). We carried out frequency‐dependent permittivity measurements for a large variety of artificial and natural substrates with an impedance analyser at frequencies ranging from 10 MHz to 1 GHz. Based on these measurements, we selected 200 MHz as the most commonly used frequency value for geological, ground‐penetrating radar applications and developed a set of empirical equations from third‐degree and ternary factorial diagrams to quantify the relationship between dielectric permittivity and volumetric moisture content of pure sand, clay, carbonate, and mixtures. The results show systematic trends, which can be explained by competing electrophysical properties of minerals and their wetting behaviour. This investigation shows that different wetting affinities of 1:1 and 2:1 clay minerals strongly control permittivity. Thresholds and non‐linearity effects were identified, where specific mineral properties and/or wetting processes became dominant in complex mixtures of soil materials. Bulk densities and porosities, respectively, were shown to have only minor influence compared with the impact of water. Consequently, the presented method is applicable to both undisturbed and disordered samples and, hence, to mobile devices or permanent installations in the field after compositional laboratory or onsite analyses. The permittivity of natural soils was measured in two field studies. The results show a good correlation between calculated and measured moisture contents. Our study confirms that the widely used Topp equation is only valid for sandy soil materials, i.e., the soil types it was originally designed for. For soils with a clay content >30% and/or a moisture level >20%, the new equations reduce the error in the level of permittivity measurements from up to 20% (using the Topp equation) to 5%.