In this paper, following the common practice, the shapes of soil particles size distributions are quantified by higher-order statistical parameters: skewness, flatness, superskewness, superflatness, hyperskewness and hyperflatness factor. Based on the experimental data from six independent experiments, precise non-linear relationships between these higher-order statistical parameters are evidenced and adequate formulas are developed by least-square fitting. These expressions could be useful in modelling and controlling the tillage quality and the resulting size distributions of soil aggregates. In opposite to the classical methods, most commonly based on histograms and mean values of the soil fraction sizes, etc., this approach is more sensitive with respect to the soil structure irregularities and it enables additional advanced check of tillage quality.