As an initial step toward a comprehensive fuel management procedure, a model was developed to optimize fuel loading patterns in pressurized water reactors (PWR) with regard to a specified function of burnup and power. For this purpose, the JUMBO code employs a simplified semianalytic function to represent X-Y power distribution and a direct search scheme to optimize shuffling. The number of feasible patterns is greatly reduced through the use of a logical set of shuffling rules. The applicability of the method was demonstrated for a typical 500-MW(e) PWR, considering both an initial and reload core.