Abstarct: Amplitude Variation with Offset (AVO) is one of the most important methods of hydrocarbon exploration that has been widely used by geophysicists in recent decades. The basic equations of AVO, which investigate how changes in the reflection coefficients of seismic waves in terms of offset, are different approximations of the Zoeprit’s equations. By generating the AVO attributes sections, it is possible to find out the pore fluid content and lithology. This method becomes more important because AVO analysis is a good method for separation of gas sand zones from wet sand zones. The seismic source wavelet is a band limited signal and composed of different frequencies. Due to the anelastic absorption during the wave propagation through the earth, the frequencies of seismic source wavelet are attenuated. The attenuation rate at high frequencies is much higher than at low frequencies, because earth acts like as a lowpass filter. Since AVO analysis requires the true amplitude, the attenuated amplitude of high frequencies in seismic data can blurry the AVO attributes sections. It has also been proven that the most effects of amplitude of seismic wave on offset are on frequencies smaller than the dominant frequency. In this dissertation, by performing AVO analysis on iso-frequency section at low frequencies and extracting attributes sections from them, an attempt has been made to solve the mentioned problem. The investigation performed on the real and the synthetic data and comparison the obtained results with the conventional AVO analysis showed that attribute sections obtained from AVO analysis in data with low frequency content have better resolution than attribute sections of analysis from AVO analysis in data with full frequency bandwidth. In the end by comparing all attribute sections obtained from both real and the synthetic data a frequency in the low frequency range was obtained which we called the "golden frequency" because its single sections will give the highest quality, resolution, accuracy, clarity of the attribute sections that are extracted in AVO analysis.