One of the current challenges in today's society is the waste management policy that arises in all spheres of society, from households to industrial operations. Gasification is one of the options that appears to be a good alternative to traditional landfilling or direct incineration of waste. In addition to components with a high calorific value such as plastics, paper and other materials, waste also contain various contaminants, which subsequently occur in the product gas. The aim of this work was to verify the effectiveness of experimental dry and wet method in the purification of gas produced by thermal catalytic gasification of RDF and to compare them with the model created in the Aspen Plus environment. The most common contaminants in waste are substances containing sulfur and chlorine. For these reasons, several experiments were realized to determine their distribution in gasification products, and it was experimentally proven that most of the sulfur and chlorine passes into a solid product. The use of Ni/Activated carbon catalyst reduced the tar content in the produced gas by 76 %. To remove S and Cl compounds dry gas cleaning by solid reactants, sodium bicarbonate and zinc oxide were tested. In the case of NaHCO3, in addition to the cleaning ability of this substance in reducing the chlorine content in the gas, there was also a catalytic effect in the cracking of tar components. By using 100 g of catalyst and 100 g of NaHCO3 at RDF feed rate 0.7 kg/h, the tar content was reduced by 87 % and the chlorine yield in the gas decreased from 0,27 mg/l to 0,13 mg/l in the best case. The same decrease occurred in the concentration of sulfur components from 435 ppm H2S and 26 ppm COS to 292 ppm H2S and 10 ppm COS. Compared to results obtained by simulation using thermodynamic equilibrium model in Aspen Plus, the real efficiency of experimental device in removing contaminants reached 87, 38, and 45 % for tar, S and Cl removal, respectively. In the case of wet methods, a biodiesel solution was used as a tar absorbent, a weak NaOH solution was applied to eliminate chlorides, and a monoethanolamine solution was added to reduce the content of sulfur components in the gas. The process of cleaning using absorbents had an inconstant character due to technological limitations and the lower selectivity of some solvents. The assembled model reliably predicts the conversion of the feed. In the case of the composition of the resulting gas, the ideal thermodynamic model provided significantly different results from the measured data.