The problem of improving the reliability of complex major equipment at Russia’s NPPs can be managed by assessing the technical condition of equipment using new generation diagnosis systems. This requires development of algorithms for diagnosing technical objects (TO) and preparation of personnel for performing diagnostic measurements and arriving at diagnosis of the technical condition of TO. The objects of diagnosis in new NPPs include a reactor, reactor coolant pumps, turbine-generator, motor- operated valves, and other systems and components. The paper addresses the issues of preparing standards and specifications for Technical Diagnostics and using unified terminology in developing diagnosis systems for TO. A special attention is made to definition of the term “diagnostic indicator” which can be understood as a structure parameter or characteristic of a TO. In this paper, diagnosis objects are considered as “continuous” objects. The “diagnostic model” notion is introduced. A diagnostic model is constructed based on a mathematical model of the TO and relates direct measurements to diagnostic indicators of the condition of the object. The authors also propose using special algorithms in the diagnosis procedure for the cases when the amount of measured information is insufficient or the structure of a diagnosis object is unknown. In the first case, integrated diagnostic indicators should be used and in the second case, mathematical model coefficients should be identified and their values monitored during the object lifetime. The proposed methodology of developing diagnostic software supports Condition-Based Maintenance (CBM) strategy which can advantageously replace the currently used Scheduled Maintenance (SM) approach.