After a historical survey retracing the evolution of maintenance practice, the concept of prognostics and health management (PHM) is introduced. Differences and similarities between PHM and reliability, availability, and maintainability, as well as links between the two disciplines, are emphasized. In contrast to traditional reliability analysis, which relies on time-to-failure data to estimate the lifetime distributions and to evaluate the reliability of the “average” asset operated under “average” operating conditions, the PHM approaches enable the assessment of the reliability an individual asset under its actual use conditions and its actual system health state. Affordable advanced sensors and measuring devices, communication networks, and computer processing power nowadays make such approaches practical. PHM standards, such as those from the International Organization for Standardization and the Institute of Electrical and Electronics Engineers, are reviewed. Performance indicators are described for the successive phases, including detection, diagnostics, and prognostics. An overview of the data integration and human–machine interface issues is given. The case is made for PHM in the railway industry. Finally, the authors describe three case studies based on their industrial experience with Alstom’s HealthHub™ predictive maintenance program: turnouts, bogies, and infrastructure (track and catenary). The chapter concludes with the information technology aspects of PHM and the authors’ views on current challenges and future perspectives.