To address the challenges posed by the complex micro-nano pore structure and strong heterogeneity of shale reservoirs, which limit the accuracy of conventional permeability-based damage evaluation methods, a new evaluation approach based on nuclear magnetic resonance (NMR) combined with T₂ spectrum normalization was proposed. Shale cores from the Chang 7 Member of the Yanchang Formation and the Shanxi Formation in the Ordos Basin were selected as representatives. Simulated damage experiments were conducted using typical fracturing fluid systems, with NMR signals collected online from small-sized cores and analyzed through normalized processing. The results were then compared with traditional permeability-based evaluations for validation. The results indicate that the two types of reservoirs exhibit significant differences in their response to fracturing fluids: in the Chang 7 reservoir, the clean fracturing fluid system shows the highest normalized fluid intake rate change (45.06%), while the clean water system has the lowest (-3.19%). In the Shanxi Formation reservoir, the small-molecule system performs best (41.16%), whereas the guar gum system shows relatively poorer performance (5.02%). The damage trends of each system in both reservoir types are highly consistent with the permeability method, validating the accuracy and reliability of this approach. The findings demonstrate that the “online NMR+ normalization analysis” can effectively reduce the influence of core heterogeneity and size differences, enhancing the comparability and stability of evaluation results. This method provides a rapid and robust tool for shale reservoir damage assessment and fracturing fluid optimization.