In this paper, a long-term Distribution System Planning Model (DSPM) considering network reconfiguration and Distributed Generation (DG) integration, is presented. The DSPM objective functions are power loss reduction and voltage stability enhancement. The proposed model handles long-term load increase and seeks to define when and where the reinforcements, such as network reconfiguration (NR) and DG integration, are required to meet the load increase. A Binary Particle Swarm Optimization algorithm (BPSO) is used for network reconfiguration while optimal DG allocation and sizing are performed using the Voltage Stability Margin Index (VSMI) and curve fitting approximation. The planning horizon of each solution is determined through load forecasting using a Nonlinear Autoregressive neuronal network with exogenous variables (NARX) technique. The performance of the proposed model is evaluated considering several case studies on the IEEE 33-bus test system. Results demonstrate the effectiveness of the proposed approach.