This paper presents results of a study of dielectric properties of snow water with low solutes concentrations and of water from strongly mineralized flat land lakes of Altai region. Experimental data are processed with multivariate regression analysis projective methods, and regression dependence of dielectric properties on mass concentrations of chemical substances and measured frequency for multicomponent solutions is elaborated. In the paper, regression dependencies of real and imaginary parts of the complex permittivity on mass concentrations of chemicals in solutions are presented. The study is conducted for the following chemicals: CO₃^2–, HCO₃^–, Cl^–, SO₄^2–, Ca²⁺, Mg²⁺, N0₂^–, NO₃^–, PO₄^–3, F^–, etc. Numerical values of the complex dielectric constant can be calculated with the proposed dependencies for the known values of mass concentration. Thus, the dependencies can be used in monitoring of environmental situation and state changes of natural objects.