Abstract: For offshore water flooding sandstone reservoir, it is urgent to address how to quickly and qualitatively identify the water flooding layers during the water flooding stage. The main change of logging response characteristics is resistivity in water flooding layer, then the original resistivity reconstruction becomes the key for identifying water flooding reservoir. Based on rock physics model, the water flooded layer conductive mechanism and the influencing factors are analyzed, and the sandstone reservoir volume conduction model is established. The original resistivity of the formation was calculated by the calculation formulas of bound water saturation and Indonesian saturation, and reconstructed by using the MRGC neural network method at the same time. The results of the two methods are the same with adjacent well logs in the same layer. A method for qualitative identification of water flooded zones using the envelope amplitude difference between the reconstructed curve and the current resistivity curve is proposed, and a qualitative identification chart for water-flooded zones is established. The identification chart of water flooded layers can qualitatively indicate the water flooded situation of reservoirs, and provide a fast and reliable basis for oilfield water control and oil production increase as well as development adjustment.