A significant proportion of the UK energy needs are currently serviced by a fleet of ageing nuclear reactors. Ensuring that these reactors are operated safely is the highest priority and the structural health of their cores, that provide channels for control rods and coolant gas, is a key aspect. This thesis focuses on the application of structuralhealth monitoring to the graphite reactor cores used in the UK and presents a specification for the use of structural health monitoring (SHM) techniques already es- tablished in bridge and aircraft monitoring, with data obtained through existing reactor monitoring processes. This approach utilises statistical and clustering techniques on monitoring data that can be acquired during online operation of the plant. The use of existing monitoring processes to complement the established inspection regime for nuclear reactors is a novel contribution from this work. As part of proving the SHM approach, this thesis reports on work undertaken to identify suitable data and numerical limits for the cluster analysis. This analysis considers the data with respect to the stated aim of detectin~ core distortion and demonstrates that the chosen data and values are acceptable and conservative in the context of reactor condition monitoring. An assessment of the SHM solution is presented describing the im- plementation of the SHM approach using a multi-agent system (MAS), IMAPS. This implementation required consideration of using MAS tech- nology for condition monitoring, and the novel contribution of a technique for storing and retrieving historical data in a manner concomitant with both MAS and relational database theory is presented.ij The thesis concludes that condition monitoring is feasible on the graphite cores, and that multi-variate analysis through SHM implemented within a MAS offers a storage and analysis platform that can both handle the data volumes and accommodate further extensions as required.