Railway sleepers represent an essential element of the track; indeed, their structural integrity is closely related to important technical and safety issues. Today, periodical visual inspections are the only method applied to check the status of sleepers but are limited to visible surfaces, whereas the early detection of in-service cracks in the whole volume of sleepers would provide great advantages in terms of maintenance and management. The aim of the paper is to propose an acoustic emission (AE)-based structural health monitoring (SHM) approach that is able to detect the initiation and propagation of cracks in in-service pre-stressed concrete sleepers. The investigation is carried out in the laboratory, comparing the results obtained by acoustic emission monitoring and digital image correlation when subjecting pre-stressed concrete sleepers, taken from production, to both static and cyclic loads. The main points dealt with in the paper include the sensitivity of acoustic emission to detect damage development and the signal processing approach needed for defining effective damage indexes. Given the encouraging results, the paper is the first step in developing an affordable monitoring system, to be embedded into sleepers, that is able to be part of a complete track monitoring system.