The effect of N fertilization on yield and TSNA content of Burley tobacco was evaluated in 2007 and 2008 at two sites in Tennessee, one in southwestern Virginia, and one in central Kentucky. An additional site in southern Illinois was included in 2007. The 2007 studies evaluated all combinations of five preplant N rates (0, 67, 135, 202 and 270 kg/ha) and two sidedress rates (0 and 67 kg/ha) on the variety KT204LC. The 2008 studies included four varieties: KT 204LC, TN 90LC, NC7LC, and a TN90 line selected for high conversion of nicotine to nornicotine. The highest preplant N rate was dropped in 2008, and only five N rate treatments were used with TN90 high converter and NC7LC. Chlorophyll meter and petiole nitrates were found to be of little value in evaluating response to N. A critical PSNT (presidedress soil nitrate test) level of 46 ppm nitrate N was established for yield response. Using a multiple range test, yields were not increased by N rates greater than 135 kg/ha. Using a quadratic plateau model, yields reached a maximum at 205 kg N/ha. Yields with no N fertilizer were surprisingly high, 2713 Kg/ha in 2007 and 2118 kg/ha in 2008. Most fields used in the study were in the first year of tobacco after two years of grass sod. TSNA content in 2007 increased with N rate, but the quadratic plateau model indicated a plateau at 258 kg N/ha. Optimum N rates for Burley yield across all sites were below current recommendations based on a multiple range test, and were in the middle of the recommended range for most states based on the quadratic plateau model.