In the past few years several new rapid prototyping methods have emerged. These rapid prototyping systems use various methods to create physical parts from existing CAD files, within a short amount of time. Since less time is required to produce a prototype, the final product is able to reach the market sooner. The physical prototypes can then be used to evaluate the integrity of the design. Whenever possible the prototype should be as similar as possible to the desired finished part. Fused deposition modeling, which is often referred to by its initials FDM, is a type of rapid prototyping technology commonly used within engineering design. The technology was developed by S. Scott Crump in the late 1980s and was commercialized in 1990 . The FDM technology is marketed exclusively by Stratasys Inc. FDM is one of rapid prototyping process that uses thermoplastic materials such as ABS (acrylonitrile-butadiene-styrene) in the semi molten state to produce prototypes. FDM is an additive process and the prototypes are made by layer by layer addition of the semi molten plastic material onto a platform from bottom to top. The quality of FDM produced parts is significantly affected by various parameters used in the process. This dissertation work aims to study the effect of three process parameters such as layer thickness, sample orientation and raster angle o property of FDM processed parts. Process parameters such as layer thickness, raster angle and part orientation in addition to their interactions are studied in the present dissertation that influences the Impact resistance, Fatigue resistance and dimensional accuracy of the part produced by the process of FDM. Due to shrinkage of the filaments, the dimensions of the CAD model does not match with the FDM processed part. Influence of each parameter on responses such as impact resistance, fatigue resistance and percentage change in length, width, and thickness of the build part are essentially studied. The effect of process parameters on responses are studied via Response surface methodology (RSM). RSM is used to calculate the regression coefficients and the function is made with the significant factors. In order to reduce experimental runs, due to time saving and material cost, response surface methodology based on box-behnken design is adopted. Specimens are prepared for test as per ISO standards and modeled in CATIA V5 software. Specimens per experimental run are fabricated using Rapman 3.2 FDM machine. Empirical relations among responses and process parameters are determined and their validity is proved using analysis of variance (ANOVA) and the normal probability plot of residuals. MINITAB16 software is used for statistical analysis. As FDM process involves large number of conflicting factors and complex phenomena for part building, it is difficult to predict the output characteristics accurately by conventional method. So, an artificial neural network with back propagation algorithm has been adapted to model FDM process. Then optimization of process parameters is made by ge netic algorithm so as to maximize impact resistance, fatigue resistance and minimized percentage change in dimensional of specimen. Non dominated sorting genetic algorithm (NSGA-II) is used to optimize three responses together. MATLAB2011 software, is used for implementation of ANN and GA algorithm. Keywords: Rapid prototyping, FDM, Genetic algorithm, NSGA-II, ANOVA, ANN