A new approach to steady state modelling of vapour-compression liquid chillers is presented in this paper. The model uses a generalised radial basis function (GRBF) neural network to predict chiller performance. The GRBF chiller model is developed with the objective of requiring only those input parameters that are readily known to the operating engineer, i.e. the chilled water outlet temperature from the evaporator, the cooling water inlet temperature to the condenser, and the evaporator capacity. The GRBF chiller model predicts relevant performance parameters of a chiller, especially the coefficient of performance. The neural network is applied to two different chillers operating at the University of Auckland, New Zealand and the agreement is found to be within ±5%. It is inferred that neural networks, in particular the generalised radial basis function, can be a promising tool for predicting the chiller’s performance for fault detection and other diagnosis purposes.