In recent years, Additive Manufacturing (AM) has been attracting attentions. AM allows us to fabricate products that have complex shapes, including ones with those that are difficult to be shaped by traditional processing techniques such as forging, cutting, and plastic work [1]. However, the heating process induces a residual stress and deformation in the product [4]. It is important to predict how residual stress and deformation arises, to improve the safety and the accuracy of dimension of the products. In this study, an analysis method using a large scale parallel computation with the finite element method is presented for AM. Investigations on the effect of parameters such as the number of layers and the direction of heating on the distribution of residual stress and deformation are carried out using a simple shaped block as an example by performing heat conduction analysis and thermal elastic-plastic analysis. Hierarchical Domain Decomposition Method (HDDM) [2][3] is applied to the analyses as the solution techniques. To predict how the thermal strain arise, it is needed to perform heat conduction analysis first. In thermal elastic-plastic analysis, we can perform the structural mechanics analysis using the temperature history gained by heat conduction analysis as the internal force. In the conference, the modeling for this problem and the result of this analysis focusing on the effect of number of layers and the laser scan pattern on deformation and residual stress as illustrated in Fig. 1 will be presented.