This paper deals with the Distributed Generation (DG) capacity of the electrical power system. DG capacity is formulated as a Mixed Integer Programming (MIP) problem with respect to various operating constraints in power generation. The objective of this model is to minimize the real power loss by increasing the system voltage profile. DG is used to reduce the line losses, improving system voltage profile. The optimum value of this model is obtained by Lagrangian Relaxation (LR) technique. Lagrangian function gives the optimal value of the DG capacity which is connected in the existing system. The proposed methodology is compared with the other optimization technique Particle Swarm Optimization (PSO) and the results tested through IEEE-30 bus system. Based on numerical calculations and graphical representation, the optimum value of DG is obtained by increasing the maximum loadability of the system.