In what amounted to a kind of census of sweet corn grown for processing, three years of data from 175 fields in Illinois, Wisconsin, and Minnesota shed light on what works and what doesn’t.