An application of the feed-forward neural network for simulation of the absorption column has been presented. Analytical results based on the rigorous modeling of the column have been applied as a training data. The absorption column is modeled as a system with 14 inputs and 10 outputs. A good agreement between neural and analytical model has been obtained. The obtained results show that the neural network can be used to simulate the absorption column performance. This neural network model can also be used for another purposes, for instance, for steady state control of the column.