A general critical heat flux (CHF) prediction method with a wide applicable range and reasonable accuracy is essential to the thermal-hydraulic design and safety analysis at the conceptual design stage for a new pressurized water reactor (PWR). In this study, the Korea Advanced Institute of Science and Technology (KAIST) liquid sub-layer dryout CHF prediction model for Departure from Nucleate Boiling (DNB) region has been implemented in a sub-channel analysis code. and investigated for the method's possible use in a rod bundle environment with various non-uniform axial power shapes. The KAIST model showed comparable prediction capability to Lin's method for bottom-. center-. and top-peaked heat flux shapes. The KAIST model. without any correction factors or empirical constants, turned out to be suitable to fulfill the needs for a basis of a general CHF prediction method as compared to Lin's method and Westinghouse-3 (W-3) correlation. (C) 2005 Published by Elsevier B.V.