The dynamic hysteresis loops of a range of strip wound cores have been obtained over a wide frequency range (50-1000 Hz). A dynamic hysteresis model from measurements using an artificial neural network has been developed. Input parameters include the geometrical dimensions of wound cores, peak magnetic induction and magnetizing frequency. The neural network dynamic hysteresis model has also been compared with the dynamic Preisach, energetic and viscous-type dynamic hysteresis models after they have been applied to toroidal wound cores. The results show the neural network model has an acceptable estimation capability for dynamic hysteresis loops of toroidal cores. (c) 2005 Elsevier B.V. All rights reserved.