Pipeline transportation of resources is considered a vital method due to low operational cost, and simple design and implementation. However, potential leaks compromise the integrity of this method. Pipeline leaks consequences are major concerns due to resources loss, environmental impact and potential human injuries and fatalities. This paper investigates neural network based probabilistic decision support system for detecting the presence of leak in pipeline transportation systems. The probabilistic model correlates measurements of inlet and outlet pressures and flow to leak status. Several pipeline leak detection methods have been developed, nevertheless, noisy data, and changes in operational conditions are the main challenges that limit the performance of leak detection leading to high false alarms. ANN properties of noise-immunity characteristics, parallel structure and correspondingly fast processing and classification capabilities provide enhanced performance of leak detection.