Abstract: As a strategic mineral resource that integrates coal-based fuel, liquid oil, and gaseous energy, the efficient exploration and tiered utilization of tar-rich coal are of great significance for ensuring national energy security and maximizing the value of coal resources. To address the current limitations in understanding the organic geochemical properties of tar-rich coal and the overreliance on tar yield as a sole evaluation criterion, this study integrates coal petrographic analysis, fractionation of soluble organic matter, and biomarker compound analysis to systematically investigate the petrographic features of coals with varying tar-bearing characteristics, the compositional profiles of soluble organic fractions, and the distribution patterns of biomarker compounds. The study further examines the correlation between tar yield and coal petrology–geochemical parameters. Results indicate that vitrinite and liptinite exhibit a significant positive correlation with tar yield, whereas inertinite shows a significant negative correlation. The chloroform asphalt “A” mass fraction (average 0.73%) in tar-rich coal (include tar-high coal) is significantly higher than that of tar-containing coal (average 0.44%), and demonstrates a strong positive correlation with tar yield. Additionally, the saturated hydrocarbon content in tar-rich coal (including tar-high coal) averages 21.99%, significantly exceeding the 10.22% observed in tar-containing coal; in contrast, the aromatic hydrocarbon mass fraction in tar-rich coal (average 25.65%) is considerably lower than that in tar-containing coal (average 41.33%). Based on these findings, a comprehensive classification system comprising three categories each for maceral composition, group components, and biomarker parameters was established. Sedimentary environment analysis reveals that tar-rich and tar-high coals formed under reducing conditions, with organic matter predominantly derived from higher plants (e.g., Cupressaceae, Pinaceae) and phytoplankton. The research findings enable a more scientific and comprehensive prediction and evaluation of the oil-bearing capacity of coal seams. They effectively address the limitations of relying solely on tar yield as an indicator, providing new quantitative metrics for predicting coal tar content. This holds significant guiding value for the tiered utilization and efficient development of coal resources.