Many owners of Pre-stressed Concrete Cylinder Pipe (PCCP) water mains around the world experienced failures in their pipelines. The condition and performance of any water pipelines can be assessed by conducting direct inspection (electromagnetic, acoustic monitoring, GPR radar, etc), which when applied to an entire pipeline can be prohibitively expensive or impossible due to operating conditions. As an alternative, it is possible to use intelligent models to predict the condition and performance assessment of the pipelines based on historical condition observations and inspections results.In this research work, a PCCP wire breaks prediction model is developed using Artificial Neural Network (ANN) technique and applied to real-world acoustic monitoring data collected from Great Man-made River Project (GMRP). The developed model involves nine input variables that affect the deterioration process of PCCP. These variables are: monitoring period, pipe age, soil resistivity, design pressure, design soil density, design soil cover, type of pre-stressing wire wrap, wire diameter and wire pitch. The output of the model is the number of wire breaks, which is considered the most important factor in evaluating the condition and performance of PCCP. The ANN model demonstrates well representation to the training patterns and shows good prediction performance ("R2 "~0.99 for groups of Special Pipes and "R2"-0.60 for groups of Standard Pipes). The results of ANN model are compared with that of Multiple Linear Regression (MLR) model ("R 2"<0.28 for both Special & Standard groups of pipes).