A methodology has been developed for determining the family of near-optimum fuel management schemes that minimize the levelized fuel cycle costs of a light water reactor over a multicycle planning horizon. Feed batch enrichments and sizes, burned batches to reinsert, and burnable poison loadings are determined for each cycle in the planning horizon. Flexibility in the methodology includes the capability to assess the economic benefits of various partially burned batch reload strategies as well as the effects of using split feed enrichments and enrichment palettes. Constraint limitations are imposed on feed enrichments, discharge burnups, moderator temperature coefficient, and cycle energy requirements. The methodology, incorporated into a code named OCEON, uses a zero-dimensional reactor physics model and a rapid fuel cycle cost routine to select minimum cost cycling schemes that satisfy all constraints. These candidate schemes are then examined with a two-dimensional nodal reactor physics model to more accurately calculate feed enrichments, batch burnups, and fuel cycle costs. The use of Monte Carlo integer programming to direct the optimization process allows for the determination of a family of low cost schemes from which the fuel manager can select the strategy that best fits his needs.