Power loss and reliability are concern of any power networks. This thesis try to address these problems by integrating distributed generation in a network. Optimization part is executed in two steps. Loss sensitivity factor is used to find the best location for distributed generation and in second step genetic algorithm optimization is used to find optimum size of distributed generation. The optimization uses minimization of active power loss and sum of voltage deviation. The reliability improvement is also studied for the base case and after allocation and rating of DG. The method is evaluated on 84 nodes 15 kV outgoing feeder 02 of Addis West substation. The feeder has fifty-five transformers with a connected load of 12 MW. For base case analysis, all bus voltages found to be within standard, 0.95 p.u - 1.05 p.u, for all load scenarios except at connected load. For connected load analysis, bus numbers 68 possess a magnitude of 0.9482 p.u. Sum of per unit voltage deviation of each bus for same load profile found to be 3.1271 p.u. Also, the highest base case active and reactive loss is calculated for connected load scenario. The values found to be 552.7 kW and 331.6 kVAr. The reliability assessment index of the feeder is far behind the Ethiopian Electric Agency standard. The agency stated System Average Interruption Frequency Index (SAIFI) of 20 Int/Yr and System Average Interruption Duration Index (SAIDI) of 25 Hr/Yr. Base case SAIFI and SAIDI of the feeder are 144.68 Int/Yr and 115.32 Hr/Yr, respectively. Following the optimization procedure, two 0.6 MW PV DGs are installed at bus 68 and 21. The result shows improvement of all the targeted physical parameters of the feeder. Bus voltages are maintained with in standard limit for all load profile scenarios. 10.52% reduction is found for the sum of per unit voltage deviation. The best reduction of 16.25% in both active and reactive loss is calculated for 50% load scenario. The improvement found for SAIFI and SAIDI with integration of DGs is 10.35% and 10.42%, respectively. Keywords: Distributed Generation, Voltage Profile, Reliability, Active Power