The objective of this paper is to present a method for determining diurnal distribution of the intensity of effective non rainfall water flux. It was found that the application of TDR technique for the determination of diurnal dynamics of effective non rainfall water flux requires temperature correction of sensed volumetric moisture contents. Without temperature correction the error of estimated non rainfall water flux can be as much as 26%. In addition, the effect of temperature changes on the soil surface was determined in 0.5, 1, 2, 3, 4, and 5 hours periods. It was found that the intensity of effective non rainfall water flux was determined to the greatest extent by the rate of temperature drop during the period of 3 h preceding the non rainfall water flux determination. The agreement of non rainfall water flux calculated with the method proposed and that obtained by the collector was better for dew than for hoarfrost periods.