Epidermal growth factor receptor monoclonal antibody was approved for treatment of metastatic colorectal cancer patients carrying KRAS wild type DNA. However, recent studies showed that patients with KRAS G13D mutation may benefit from EGFR antibody therapy. In this study we tried to explore whether the abundance of KRAS mutation could affect the efficacy of EGFR antibody therapy. We firstly established a PNA-PCR method which could calculate the percentage of KRAS mutation in total DNA and proved its ability on 47 colorectal cancer samples bearing KRAS mutations. Then we analyzed the correlation between the abundance of KRAS mutations and efficacy of EGFR antibody therapy in another 35 metastatic colorectal cancer patients. We proved that PNA-PCR assay could calculate the abundance of KRAS mutation and the percentage of mutant DNA in tumor cells varied a lot (10.8%∼98.3%) on the 47 colorectal cancer patients. The efficacy of EGFR antibody correlated with the abundance of KRAS mutations: in the KRAS mutation less than 30% group, the disease control rate was 44.4% (4/9); the disease control rate of 30∼80% group was 5.6% (1/18) and the >80% group was 12.5% (1/8) (P = 0.038). In summary, our study showed that PNA-PCR method could easily detect the percentage of KRAS mutation in tumor cells and colorectal cancer patients with low abundance of KRAS mutation might benefit from EGFR antibody therapy.