Abstract: Regarding the problems that a large number of low-contrast pay zones exist in Dongying formation of the Block Nanpu 2-1 in Jidong oilfield, Bohai Bay basin and conventional logging methods cannot be used to accurately evaluate oil-bearing properties of reservoirs, the paper introduces the natural gamma relative value and apparent formation water resistivity as valid parameters, uses logging chart to eliminate influences of lithology and water on electrical properties, applies probabilistic neural network (PNN) method to enhance logging identification accuracy, which could solve the contradiction between electrical and oil-bearing properties, and finally develops an evaluation technology taking amplifying electrical characteristics as the core. The research results have guiding significance for low-contrast pay zone identification by well-logging analysis