Rare elements in nature are often found in the form of oxide, and since they are very similar in physical and chemical properties, their separation is very difficult. Hence, the study and selection of the most appropriate methods for the separation of rare elements are very important and necessary. Therefore, in this first step, in order to model and optimize the effective factors on improving the separation of Dispersum and Gadolinium, five factors including pH, Cyanex 272 to Depa ratio, solvent extraction’ s concentration, time and lactic acid’ s concentration were selected. Then, DX7 software was used to design the experiments and based on that, 29 tests were determined. But on the other hand, in carrying out experiments, such things as weighing with the device, preparing solutions with a specific concentration, removing different volumes, and etching of the items that can cause the error and therefore the uncertainty is why the discussion of probabilistic analysis in order to Modeling under conditions of uncertainty can be very useful for the assurance of the results. For this purpose, in the second step, based on the data obtained from the experiments and the relationships obtained between the input and output variables, first, the type and characteristics of the probability distribution function, each of the five effective input parameters were determined and based on them and with Considering the simulation results, the characteristics of the distribution parameters of the output parameters (percent recovery of dysprosium and gadolinium) were obtained.