A review of worldwide results reveals that reaction rates in the blanket region are generally underpredicted with the discrepancy increasing with penetration; however, these results vary widely. Experiments in the large uniform Purdue University Fast Breeder Blanket Facility blanket yield an accurate quantification of this discrepancy. Using standard production code methods (diffusion theory with 50-group cross sections), a consistent calculated-to-experimental (C/E) drop-off is observed for various reaction rates. A 50% increase in the calculated results at the outer edge of the 51-cm blanket is necessary for agreement with experiments. The usefulness of refined group constant generation, utilizing specialized weighting spectra, and transport theory methods in correcting this discrepancy is analyzed. Refined group constants reduce the discrepancy to half that observed using the standard method. The surprising result is that transport methods have no effect on the blanket deviations; thus, the present multigroup transport theory does not constitute or even contribute to an explanation of the blanket discrepancies. The residual blanket C/E drop-off (about half the standard drop-off) using advanced methods must be caused by approximations that are applied in all current multigroup methods.