Oil fingerprinting technol. has been used in the upstream industry for decades to trace back the contributing geol. formations that the oils were produced from.Traditionally, oil fingerprinting relies on 1D GC which only resolves less than ∼100 compounds and the repeatability is limited, constraining its application in the unconventional reservoir.However, shale reservoir characterization requires much higher resolution and accuracy distinguishing geochem. variation between every ∼10-15' thickness of rock.With both hardware improvement and advanced data processing tools, we developed comprehensive GCXGC method that resolves thousands of compounds with significantly improved repeatability.Using 156 repeat samples, the P90 repeatability falls within 0.35%-1.04% while P80 falls within 0.39%-1.00%, significantly improving the robustness of technique and expanded its application.We studied 30,000+ shale crude oil samples from over 2,500 wells across all major shale reservoirs in the North and South America and results from selected samples will be presented.Unique features of these oils with different organic origins, depositional environment and migration pathways were disclosed, which are routinely used to optimize shale reservoir development in the upstream productionThe composition of crude oils also has crucial impacts on the downstream refining processes.This large-scale detailed fingerprinting study of crude oils from various formations and basins may help to customize and improve the efficiency of downstream refining processes.