The work presents a research on inventory ABC classification using various multi-criteria methods (AHP method and cluster analysis) and neural networks. For the real inventory sample data and previously conducted traditional ABC analysis the applications of the mentioned methods in inventory classification have also been investigated. The applied methods’ obtained results have been used to evaluate their usage possibilities in real manufacturing environment. The investigations carried out in the present work create real conditions for a better inventory control and implementation of the results in the ERP system inventory module