This paper presents a comparison of component-wide and plant-wide monitoring studies based upon neural networks for anomaly detection in Borssele nuclear power plant. The Borssele nuclear power plant is a unique nuclear facility to produce the electricity through the nuclear energy in the Netherlands. This power plant was established as a Siemens-KWU design and started to operate in the commercial sense in 1973. However, it was modified under the backfitting programme in 1997. After the backfitting programme, its safety level was raised to the current state of the art. This paper covers the component-wide and plant-wide monitoring studies using the neural network approach for this modified case of the power plant. (c) 2008 Elsevier Ltd. All rights reserved.