A numerical method is proposed to predict the stochastic dispersion of data that unavoidably occurs in a real drying process. The method is based on the use of discrete frequency distribution curves to predict a random variation on some parameters of the process and is applied to the case of radio frequency vacuum (RFV) drying of wood. Experimental data of RFV drying of western hemlock timbers were used to obtain the numeric values of the stochastic parameters and their standard deviation. A mathematical model was designed to simulate the final moisture content dispersion on hypothetical examples. Results were quantified using three indices: average, range and standard deviation.