This paper uses data from thirty-three on-farm experiments to explore the use of electrical conductivity (EC) for defining seeding and nitrogen rates for corn production. We estimate the yield response to nitrogen and seeding rates, including an interaction term with EC for each of the trial years. We then determine the optimal uniform and variables rates and compare the profits. If EC can be used on different fields and years, then the correlation between EC and the optimal rates should be consistent across fields and years. We find that the optimal variables rates do not produce profits above $5 an acre for the majority of the fields. Additionally, in different years on the same field, the high EC areas may require more or less of the inputs. The inconsistency of the relationship between EC and the optimal rates does not enable EC to be accurately used for variable rate applications across different growing years. While EC will continue to be important in detecting salt affected soils and can be calibrated for detection of specific soil elements, the use of EC for variable-rate input management is not recommended.