This paper describes a novel approach of harmonics detection in a power system which can be used as an alternative to the conventional approaches. The proposed approach uses the multilayer feed forward neural network to determine the harmonic components in a six-pulse bridge converter. In this paper the detection of 5th, 7th, and 11th harmonic components from the distorted waves has been verified by means of the computer simulation. It is found that once trained by the learning algorithm, the neural network can determine each harmonic component very effectively and efficiently.