In-service rails and Hadfield steel crossings may develop structural defects due to the stresses and environmental conditions they are subjected to. Inspection and assessment of the growth of these defects in service is critical in making maintenance decisions and ensuring the smooth operation of the rail network. The main aim of this project has been to investigate the possibility of using acoustic emission (AE) technique to monitor structural integrity of rails and crossings in real time. AE technique has been used to monitor the damage evolution process of three different types of samples during low and high frequency bending fatigue crack growth tests. The sudden increase in crack length caused by the occurrence of brittle cleavage fracture has been successfully detected by AE. Various signal processing algorithms have been employed for the in-depth analysis of the acquired raw AE waveforms and proved to be effective in evaluating the severity of damage. Overall, this research has shown that AE can be used as a tool to effectively evaluate the structural integrity of rails and Hadfield steel crossings online.