In the paper a stochastic model of the batch sieving procedure is presen- ted resulting in a system of ordinary differential equations to describe the time dependence of the mass fractions on the sieve surfaces being randomly influenced by sampling errors. The coefficients of the system can be interpreted as sieving velocities._x000D_
Their dependence on the sieving con- ditions is discussed and a special type of functional relationship and methods are suggested to determine this function experimentally. Moreover the applicability of the approach is demonstrated by an example to evaluate the particle size distribution, if only the result of incomplete test sieving is known. For this problem a rough error estimation is performed.