Abstract: gneous rock develops in shallow and middle layers in Nanpu oilfield. The igneous rock is characterized by complex distribution and fast alternation of lithofacies and thickness, which affects the hydrocarbon accumulation scale and its recognition as a whole. Drilling formation indicates that the igneous rock is mainly of basalt and tuff. On seismic profiles, the characteristics of basalt are similar to that of glutinite at the bottom of Guantao formation. In conventional wave impedance inversion profile, the impedance value of tuff is similar to that of sedimentary rock. Therefore, the integrated analysis of logging data was carried out firstly and the pseudo-lithology curve was established through logging lithology and sensitive curve, which can reflect the difference of lithology. Then, the pre-stack AVO processing was used to obtain attribute data volume, including gradient (G), intercept (P), and Poisson ratio (σ) to extract seismic attributes including amplitude, frequency and phase. Finally, taking the pseudo-lithology curve as the target one, the seismic attributes with high sensitivity was optimized as inversion parameters, PNN inversion technology based on neural network was used to identify and describe the igneous rock in Nanpu oilfield. The result coincides with the geological recognitions.