Current and voltage waveforms of a distribution or a transmission system are not pure sinusoids. There are distortions in these waveforms that consist of a combination of the fundamental frequency, harmonics and high frequency transients. This paper presents an enhanced measurement scheme for identification and tracking of harmonics in power system. The proposed technique is not limited to stationary waveforms, but can also estimate harmonics in waveforms with time-varying amplitudes. This paper presents a new method based on Real Coded Genetic Algorithm, which is a technique for optimization inspired by genetics and natural evolution. The algorithm was tested using simulated data, and effects of sampling rate studied. Results are reported and discussed.