The machining technologies have always been facing different challenges such as machining hard and brittle materials, getting better surface quality, getting required dimensional and geometrical tolerances, reducing machining forces, increasing tool life, reducing the burr size and etc. Studies in different fields are being done to fulfill the industry needs, improving the cutting tools, machineries and optimization of machining parameters. Such as Alloy steel AISI 4140 is steel with low amount of molybdenum which is used in manufacturing of industrial parts, especially in air and chemical industries. Machining of this steel is very difficult because of existence of nickel and molybdenum in this kind of steel alloy.GA is a heuristic optimization algorithm that finds the optimal solution quickly when the analytical or empirical model is available. The neural network model coupled with GA is proposed to represent the relationship between the cutting conditions and the cutting-related variables, using neural networks, and to determine the optimal machining parameters, using the genetic algorithm with minimal human interference. In view of the number of factors and continuous range of values, a strategy for reduction number of experiments should be devised. Taguchi’s DOE method was thought to be appropriate for that purpose. The Taguchi method is an experimental design technique, which is useful to accommodate this purpose by using a system of factors and their levels that is called orthogonal arrays. In this study, an artificial neural network (ANN) model based on experimental measurement data was developed to predict surface roughness and cutting force components in face milling of AISI 4140 steel. In order to attain minimum operation numbers and decrease the cost of machining, an experimental scheme was arranged taking advantage of Taguchi method. The considered parameters were cutting speed, feed, depth of cut and engagement. Back propagation artificial neural network was utilized to create predictive models of surface roughness and cutting forces exploiting experimental data, and the GA algorithm was used to find the optimum of surface roughness. Cutting force components and surface roughness were measured, and then analysis of variance (ANOVA) is performed. Based on the experimental results presented, the following conclusions can be drawn from face milling of AISI 4140 steel. Surface roughness increases when the feed increases, and surface roughness decreases when the cutting speed increases; with a higher cutting speed and a lower feed, it is possible to obtain a better surface finish. Cutting forces increase when the feed increases, and increase in cutting speed leads to decrease in Cutting forces; i.e. with a higher cutting speed and a lower feed, it is possible to obtain lower cutting forces. Back propagation artificial neural networks can be employed reliably, successfully and accurately for the modeling of surface roughness, cutting forces and prediction of their values in face milling of AISI 4140 steel. Finally in order to validate the method, an experiment with the obtained optimal cutting condition was carried out, and the results were compared with the predicted value of surface roughness. The corresponding results show the capability of GO to predict surface roughness. Key words AISI 4140 alloy steel, variance analysis, neural network, genetic algorithm.