The control and operation of complex power-generating systems, such as nuclear power plants, rely on the measurements of several sensors that monitor the process and the system state. On the basis of the sensor measurements, the system is operated for maximum economic efficiency and safety. Out-of-calibration sensors can lead to misinterpretation of the system state and problems with control and operation of the process, with possible economic losses, equipment damage, and safety consequences. To avoid such occurrences, periodic sensor calibrations are scheduled to ensure that sensors are operating correctly. These calibrations are performed manually and involve all sensors, independent of the actual need for calibration of each sensor. Continuous sensor calibration monitoring would then be most desirable both to ensure correct process control and system operation and to reduce maintenance costs associated with performing unnecessary manual sensor calibrations. This latter issue is of great relevance in nuclear power plants due to the large number of sensors employed, which are tested for calibration at each refueling outage. In this paper, the artificial neural network-based sensor calibration monitoring system is proposed to provide continuous sensor status information and virtual estimates for faulty sensors. In particular, we illustrate the design of an autoassociative artificial neural network for sensor fault detection and validation. The efficiency of the proposed method is verified through its application to eight critical transient signals coming from a U-tube steam generator of a pressurized water reactor modeled by means of a validated simulation code.