Identifying causal driver mutations among the vast majority of other passenger mutations is key to understanding tumorigenesis and the development of cancer therapy. Here, we developed a novel high-throughput method using cytosine base editor (CBE) and adenine base editor (ABE) to evaluate the function of somatic cancer-associated mutations found in human cancer tissues. We designed a total of 83,731 and 23,613 sgRNAs for CBE and ABE, respectively, that would lead to the generation of 107,982 single nucleotide variants found in human somatic cancers. We performed screening to identify mutations with positive (outgrowing) or negative (depleting) effects on proliferation and survival followed by two sets of high-coverage small focused library evaluation with sgRNA libraries. We found that unique molecular identifier (UMI)-based analysis outperforms conventional one-sgRNA-one-mutation based analysis in identifying outgrowing or depleting mutations. Our screening platform using base editors should facilitate cancer genomics by identifying functional consequences of individual variants, which may contribute to the development of new therapeutic options.