Deep drawing process is a flexible and inexpensive process which can be use for mass production of parts. Ductility of sheet metal deep drawing process is determined by the traction sheet and process parameters which influences the properties of deep drawing. Determination of forming limit is considered one of the most important difficultly in this process. Limit ratio the diameter of the plate (D max ) to the diameter cup (d) without breaking or tearing is called limiting drawing ratio. The relationship between the tensile properties of a metal sheet material and process parameters, is essential. This is due to the fact that the larger the ratio is, the higher will be the height of the cup. The numerical solution of the problem and spending much time is needed for each specific problem. Drawing ratio in the deep drawing process based on experimental data and finite element simulation, is predicted by a neural network. the goal for using neural network is to organize the data obtained from finite element .In this study whit organize data obtained from finite element analysis process in a neural network, produced a mathematical function, and an easier and faster way to solve this problem. Since the limit drawing ratio is effected by various parameters such as: Thickness, friction coefficient, the profile of radius die and punch, the blank holder force and the punch speed, are considered for study. The listed parameters have been combined by the L 27 Taguchi array in different levels and deep-drawing has simulated 27 different cases and the limit ratio of drawing is obtained by trial and error method for each modes. the simulation by finite element software ABAQUS is used for conducted experiments validation. Forming Limit Diagram of criteria for predicting damage initiation and growth of damage during the process and therefore plate ductility can be determined. The simulation results of used to train the network. Model of neural network has been by MATLAB program. Therefore, data obtained from similar organizing process, a neural network is an easy and fast way to solve the much difficult problem in deep drawing process for rectangular pieces and thus the relationship between limiting drawing ratio and the parameters different processes are determined. Finally, it can be seen that neural networks can be a good way for new modes predict the level of tension. Keywords : Limiting drawing ratio, Arti?cial neural network, Finite element method Deep drawing,