Abstract: Ming Huazhen Formation is the main oil-bearing series in Oilfield in Bohai Huang Hekou sag.The oil and water sands of Bohai Huanghekou sag showed strong amplitude highlight characteristics, and difficult to distinguish from poststack and prestack attributes.This paper takes the A oilfield as an example and according to the seismic response characteristics of oil and gas in this area, the low frequency area of seismic data is taken as the maximum difference frequency of oil and gas, and the frequency domain is extracted.In the time domain, amplitude exponential weighted is applied to extracted seismic data, so the oil layer is characterized by bright spot characteristic.The bright spot directly indicates oil and gas, and then reaches the purpose of hydrocarbon detection.All the sand bodies within the target area were statistically analysed.The prediction accuracy of hydrocarbon bearing sand bodies in the shallow layer of A oilfield is increased from 69% to 86% with the bright spot attribute of hydrocarbon.In particular, the prediction accuracy of the water bearing sand bodies in the oilfield is increased from 63% to 96%,and solved the problem of oil water distinction and oil gas reservoir identification in A oilfield.This method not only improves the computational efficiency, but also solves the problem of the identification of oil and gas reservoirs in A oil field.