Variants of uncertain significance (VUS) hamper the clinical application of genetic information. For example, in treating lung cancer with tyrosine kinase inhibitors (TKIs), many epidermal growth factor receptor (EGFR) variants remain classified as VUS with respect to TKI sensitivity1,2. Such incomplete resistance profiles hinder clinicians from selecting optimal therapeutic agents3,4. A high-throughput approach that can evaluate the functional effects of single nucleotide variants (SNVs) could reduce the number of VUS. Here we introduce SynPrime, a method based on prime editing that enabled the generation and functional evaluation of 2,476 SNVs in theEGFRgene, including 99% of all possible variants in the canonical tyrosine kinase domain (exons 18 to 21). We determined resistance profiles of 95% (= 1,726/1,817) of all possible EGFR protein variants encoded in the whole tyrosine kinase domain (exons 18 to 24) against afatinib, osimertinib, and osimertinib in the presence of the co-occurring mutation T790M, in PC-9 cells. SynPrime, which uses direct sequencing of endogenous regions to identify SNVs, provided more accurate functional evaluations than a guide RNA abundance-based approach. Our study has the potential to substantially improve the precision of therapeutic choices in clinical settings and contribute to addressing the issue of VUS by being applied to other genes.