Scientific classification and grading evaluation provide a theoretical foundation and key prerequisite for the cost-effective development of lacustrine shale oil in China. Based on a systematic literature review, along with comprehensive investigation and judgement, we present a thorough summary of advances in research on the classification and grading evaluation of lacustrine shale oil in China, andidentify core challenges and future trends. The evolutionary path and inherent logic of the shale oil classification system are clarified. Furthermore, we elaborate high dynamism and type-dependence of key parameter thresholds, and systematically compare the applicability and limitations of various evaluation methods ranging from multi-parameter overlay to machine learning, while focusing on the “four-property” parameter system consisting of reservoir properties, oil-bearing capacity, oil mobility, and fracability for shale oil evaluation. The analytical results reveal three major limitations in the current classification and grading evaluation. First, the classification schemes, in spite of diversity, lack a unified framework, leading to conceptual confusion and hindered popularization. Second, the evaluation parameters prove cumbersome and lack universal criteria, resulting in key parameter threshold significantly varying by basin and type, thus the absence of a unified understanding of the formation mechanisms of shale oil. Third, the disconnection between assessment methods with exploration and exploitation practices is manifested in the mismatches between geological and engineering sweet spots and broken links in the upgrade chain from resource volume to productivity. Currently, the lacustrine shale oil evaluation paradigm in China is shifting from static description to dynamic prediction and from a single geological dimension to a geological-engineering-economic multi-dimension synergy. Future efforts should be put into constructing a dynamic, full-lifecycle classification and grading assessment system by deepening the integration and applications of multi-source data and artificial intelligence (AI) technologies. Furthermore, it is necessary to thoroughly analyze typical calibration areas and develop industry standards, thus enhancing the standardization, precision, and cost-effective applications of the classification and grading evaluation of lacustrine shale oil in China.