Background: Lynch syndrome significantly increases the risk of developing colorectal cancer (CRC) due to an inherited defect in mismatch repair (MMR). Early detection relies on identification of pathogenic mutations in patients, but few canonical Lynch mutations exist. We describe two separate colon cancer patients with identical mutations in the MLH1 gene ( MLH1 c.2054C>T) and later determined to be cousins as well as a third relative diagnosed with several cancers with the same mutation. Despite a strong family history of cancer, the MLH1 mutation was labeled discordantly on different NGS panels and required several weeks to obtain results delaying patient care. This highlights a need for improved diagnostics with faster turn-around time, increased cost effectiveness, and ability to screen for non-canonical Lynch variants. We designed a novel, customizable digital PCR (dPCR) assay to rapidly detect MLH1 gene variants. We also conducted in-depth molecular modeling, mutational signature analyses, and Saccharomyces cerevisiae based functional assays to demonstrate that the MLH1 mutation likely disrupts the interaction with binding partner PMS2, impairing MMR; defining a novel MLH1 mutation. Methods: NGS: DNA was isolated. The NovaSeq platform was used for sequencing. Data analysis was performed on the RTA software. Digital PCR: TaqMan probes were created for MLH1 using a FAM labeled probe for wild type alleles and a HEX labeled probe for mutants. Quantification of fluorescent signal used Thermo Quantstudio software with total signal converted to a percentage to determine the VAF. Yeast assay: In S. cerevisiae , mlh1 and pms1 knockout ( mlh1 Δ and pms1 Δ ) strains were obtained through the fragment insertion technique. Yeast expression plasmids encoding the human versions of mlh1-2054 , MLH1 , and PMS2 were obtained via the LR clonase technique. Plasmids were transformed into the heterozygous diploid knockout yeast strain to make the genetic construct for the assay CHT8 and CHT9 in which fresh segregants are dissected into haploids for our 5-FOA assays. Results: We designed two fluorescently labelled probes, one that recognizes the wild-type MLH1 allele and one specific for c.2054C>T ( mlh1-2054). We can detect MLH1 and mlh1-2054 alleles in equal amounts, consistent with a heterozygous germline mutation in all 3 patients. We observed a significant increase in mutation frequency in yeast transformed with the mlh1-2054 variant compared with wild-type hMLH1 , which reflected the difference in mutation frequency we observed in comparing MLH1 PMS1 yeast and mlh1 Δ pms1 Δ yeast. This suggests the mlh1-2054 variant present in all patients is pathogenic due to the resulting MMR defect we see in our yeast genetic model. Conclusion: This study emphasizes the need for improved diagnostic tools to identify pathogenic mutations in diverse populations and establishes a novel MLH1 hereditary mutation.