Narrow and small channel sand bodies, characterized by their “narrow, thin, and poor” properties, are considered important targets for tapping remaining oil in fluvial-deltaic sedimentary reservoirs. Conventional prediction methods exhibit limitations: seismic attribute analysis is restricted by seismic data quality, resulting in limited identification accuracy, while reservoir inversion is generally employed to predict sand thickness and lacks specificity for identifying channel-type sand bodies. A dual-parameter “thickness-amplitude” quantitative identification method was innovatively proposed. Based on the characteristics that narrow channel sand bodies possess relatively greater thickness and higher amplitude responses on sensitive logs compared to other sand bodies within the same depositional environment, an innovative dualparameter “thickness-amplitude” identification criterion was established. On the basis of sand-thickness prediction via inversion, the amplitude attribute of the inversion volume was further extracted. Sand bodies were classified into three types: channel sand bodies, suspected channel sand bodies, and non-channel sand bodies, and narrow channels were delineated by integrating geological patterns. The core innovation lies in transforming the “morphology-amplitude” recognition pattern of well logs into the “thickness-amplitude” quantitative parameters of the inversion volume, achieving a leap from qualitative description to quantitative characterization. The method is applied to the Putaohua oil layer in a block of the Sanzhao Sag, Songliao Basin, where underwater distributary channel sands are generally 2~3 m thick and seismic data have a dominant frequency of 45 Hz, under which conditions seismic attributes fail to identify channels. Using this method, narrow channels thicker than 1.5 m are successfully identified, and a prediction accuracy of 85.2% is achieved through validation with post-drilling wells. This approach realizes “dual-parameter quantitative” identification for narrow channel sand bodies. It not only significantly improves the prediction of such sand bodies, but also provides high-precision spatial distribution constraints for remaining oil targets during late-stage field development. It can directly guide development engineering decisions such as well-pattern adjustment and selective perforation, demonstrating important practical significance for enhancing hydrocarbon recovery and extending the economic life of oilfields, with broad application prospects.