Predicting the nitrogen needs of corn for individual fields is extremely difficult.  Nitrogen recommendations based on the expected average yield response of corn or on a yield goal system with credits for previous crop and manure applications should be considered as a starting point for farmers to modify using information from individual fields.  Guidelines are needed to enable farmers to customize N recommendations for individual fields to reduce nitrate losses to water bodies like the Chesapeake Bay. One of the best methods to customize N recommendations is to establish replicated strip trials on fields with similar management to compare N rates. Twelve farmers in Lancaster County PA established a total of 31 strip trials with two or three rates of N and three or four replications from 2005 to 2010 as part of the Bay Farms On-Farm Network.  The farmers provided detailed field history information about each field.  Measurements of N status of each strip included the presidedress (PSNT) and cornstalk nitrate (CSNT) tests, and yields.  The result indicated that the CSNT and the field information, especially the manure history, current growing-season manure application, tillage, and previous crop provided useful information to fine-tune the N recommendation used as a starting point.  Reducing fertilizer applications did not significantly reduce corn yield in most of the fields and there was no significant profit loss from the reduced N fertilizer applications.  The CSNT values in the strips with reduced fertilizer applications were often in the optimal category, but the strips in fields with long-term histories of manure applications tended to have excessive CSNT values.  These data indicate that fields with long-term manure histories should be the first fields farmers target to fine-tune N management.