In this study, in order to quantify the uncertainty of the safety analysis of LBLOCA of APR1400, nuclear power plant developed in Korea, the sensitivity and correlation between variables of PIRT are analyzed and compared with the uncertainty calculated by statistical method. First, response analysis on a single variable is performed. Based on the analysis results, the importance ranking of each variable is selected and compared with the ranking suggested by PIRT. It is suggested that the importance inside of the PIRT was underestimated compared to the importance of the transition boiling phenomenon obtained in this study. In addition, through the sensitivity analysis, the relative rank of the calculated result compared to the relative change of the variable is obtained to check the importance ranking, and the important variables are found to be located at the top. Next, a response analysis on the change of two variables is performed for correlation analysis between variables. The correlation between variables was defined to observe the effect, and the importance is determined by calculating the absolute amount compared to the response of single and two variable changes. Next, the upper tolerance limit of peak cladding temperature with 95/95 condition using the non-parametric order statistics methodology is compared with other calculations. First, the peak cladding temperature limit obtained from the non-parametric order statistics methodology using only the variables in the top 10 of importance is compared with the calculation results using all variables. The results obtained from variables of high importance and those using all variables showed similar values, which means that the correlation between the variables of low importance and high importance is large. In addition, peak-to-peak intubation temperature under extreme conditions is calculated using a combination of variables and changes.