Having emerged as a non-invasive and clinically applicable approach for molecular determination of lung cancer, a genomic overview of ctDNA of large-scale cohort may be helpful in novel biomarker development and therapeutic innovation.Primary cohort encompasses 5,671 blood samples from 4,892 lung cancer patients. Pairwise tissue samples from 579 patients and additional 358 sample pairs were collected to evaluate the correlation between blood and tissue TMB. Parallel sequencing with plasma/tissue and white blood cells was performed using a 1,021-gene panel.Histological subtyping was the most relevant to ctDNA detectability independent of other demographic characteristics with small cell lung cancer showing the highest detectability, ctDNA abundance and blood tumor mutational burden (bTMB). Mutational landscape demonstrated significant differences and integrated clonality analysis highlighted distinct driver pattern and functional pathway interaction among various subtypes. The clonality and concurrent genes of EGFR mutations could predict the therapeutic efficacy of TKIs, and RB1 mutations in non-small-cell lung cancer characterized a subset with high blood TMB, elevated ctDNA level, and potential small cell transformation. Most importantly, we developed an adjusted algorithm for bTMB in samples with extremely low ctDNA level and validated its correlation with tissue TMB in an independent cohort.ctDNA could serve as a promising alternative in genomic profiling for lung cancer. The novel identification of ctDNA clonality and adjusted bTMB might improve therapeutic and prognostic evaluation. This dataset was also a valuable resource for the development of new therapeutic targets and new genomically-guided clinical trials.Copyright ©2021, American Association for Cancer Research.