Introduction: Accurate triage is important because under-triage may delay critical care for emergent patients and over-triage may inhibit efficient management of emergency department (ED) resources. In Korea, the Korean Triage and Acuity Scale (KTAS) was developed based on the CTAS in 2015. The purpose of this study was to evaluate the accuracy of KTAS in predicting patient’s severity when degree of pain was used as a modifier. Methods: This was a retrospective observational cohort study, conducted in an ED of urban tertiary university hospital with more than 90,000 visits per year. We studied adult patients who visited the ED from January 2016 to June 2016. Patients were devided into pain group and non-pain group according to whether the degree of pain was used as a modifier in the KTAS evaluation. We used acute area registration, emergency procedure, emergency operation, hospitalization, intensive care unit admission, and hospital mortality as markers to determine urgent patients. To evaluate discriminative ability of KTAS, the odds ratios of each KTAS values compared to KTAS 3 for the urgent patients were calculated. And to compare the predictive power of KTAS for urgent patients between the two groups, the area under the receiver operating characteristic (ROC) curves were compared by DeLongs method. Results: There were 9,175 (37.8%) patients in the pain group and 15,078 (62.2%) patients in the non-pain group. When KTAS was assessed as 2, only 20.3% of the patients in the pain group were registered to the acute area, while 71.2% of the patients in the non-pain group were registered to the acute area (p<0.001). And the proportion of emergency procedure, admission, ICU admission, and mortality was also higher in patients with pain group. Similarly, in the patients of KTAS 3, the proportion of urgent patients was higher in the non-pain group except emergency operation. The odds ratio for the occurrence of urgent patients decreased as the KTAS value increased in both groups, however, the difference between the odds ratios of each KTAS was more evident in the non-pain group. In pain group, compared to patients with KTAS 3, the odds ratio (95% CI) for acute area registration were 2.32 (1.92-2.80), 0.61 (0.51-0.73), and 0.35 (0.23-0.53) for patients with KTAS 2, 4, 5, respectively; in non-pain group, odds ratio were 5.59 (5.09-6.13), 0.28 (0.25-0.32), and 0.13 (0.10-0.16). The non-pain group showed better predictive power of KTAS for acute area registration than pain group; AUC (95% CI), 0.864 (0.861-0.867) vs. 0.810 (0.802-0.818), p<0.0001). The predictability of KTAS was also higher in non-pain group for emergency procedure, emergency operation, admission, and ICU admission. Conclusion: We have confirmed that the use of pain severity as a modifier in KTAS is a factor affecting accuracy. The acuity level is overestimated when pain severity is used as modifier in KTAS evaluation.