The results of soil heat flux density measurement using flux plates are as a rule burdened with substantial errors. Therefore, the verification of heat flux plate data is necessary it may be performed by comparing these data to obtained by the other method, recognised a control. In this paper, there is examined the relation between the values of soil heat flux density from plates burried at the depth of 1 cm and these obtained by combined method (i.e. heat storage changes within soil layer 1-20 cm and flux at depth 20 cm determined by gradient method). The ground for the analyses was constituted by the results of measurements carried out on the cultivated field without plants during the selected three days that were characterized by different weather and soil moisture conditions. The regression equations describing relation of soil beat flux density values (mean for one hour) derived from indirect method and flux plates in daily course were determined separately for each day. It was stated that values of the numerical coefficients in these equations differed significantly with regard to different soil moisture (as a consequence, a diverse relation of soil thermal conductivity to heat flux plate one) as well as different contact between soil material and plate. Therefore, if such equations are to be used for the correction of the flux-plates data, they should be applied, as a rule, only this day they were determined for.