Given estimates of shipping costs and scale economies for high-speed currency sorting, the authors investigate whether the Federal Reserve might lower its costs by reallocating the volume of sorting among its processing sites.