Objective To determine the positive judgement standard of K-ras mutation detection method peptide nucleic acid (PNA)-PCR/K-ras (previously established in our laboratory) and to assess its diagnostic value for colorectal cancer tissues. Methods Plasmids with K-ras codon 12 mutation and plasmids with K-ras wild-type plasmids were mixed and serially diluted into standard samples (mutation/total: 0, 1/3 200, 1/1 600, 1/800, 1/400, 1/200, 1/100 ) for six independent tests. The mutation CT, total CT and ΔCT (mutation CT-total CT) values were obtained by PNA-PCR/K-ras method. After the cut-off values of the mutation CT and ΔCT for K-ras diagnosis were identified by ROC analysis, the diagnostic criteria for K-ras mutation was defined by combining both the cut-off values of the mutation CT and ΔCT. A comparison was made between K-ras diagnostic rate by PNA-PCR/K-ras method and direct sequencing for 35 colorectal cancer tissues and their corresponding adjacent noncancerous tissues. Results The mutation CT and ΔCT values for 1/800 and the above standard samples were significantly different from those of the negative samples(P<0.05), with the optimum cut-off values of mutation CT and ΔCT being 41.7 and 15.4, respectively. The diagnostic criteria (mutation CT≤41.7 or ΔCT≤15.4) for K-ras mutation was set up as AUC-ROC 0.955 (P=0.001). According this diagnostic criteria, the K-ras mutation diagnostic rates in each concentration gradient of the standard samples (0, 1/3 200, 1/1 600, 1/800, 1/400, 1/200, and 1/100) were 0%,66.7%,83.3%,100%,100%,100%,and 100%,receptivity, with the upper diagnostic limit being 1/800. The diagnostic rates of K-ras mutation by our method and by direct sequence method for colorectal cancer tissues were 45.7%(32/70) and 18.6%(13/70), respectively, showing significant difference (P=0.000). Conclusion PNA-PCR/K-ras method has higher sensitivity and positive detection rate than direct sequencing method for colorectal cancer tissues.