Cotton producers are installing subsurface drip irrigation systems at a rapid rate in West Texas. Underground drip systems are among the most efficient irrigation systems. Nitrogen use efficiency, however, may not be as high as is commonly assumed, despite frequent injection of low rates of N fertilizer sources like urea ammonium nitrate (UAN). Our previous 15N balance studies reported only 75 % recovery of added UAN (32-0-0) in soil and cotton plant with subsurface drip irrigation. Denitrification is a likely loss pathway. To our knowledge, this report is the first on direct measurements of denitrification in any subsurface drip system. We used vented chambers and an acetylene inhibition method. Calcium carbide was inserted 12 cm in the soil over the 30-cm deep drip tape 2.5 hr before sampling the chambers at 0, 12, and 24 minutes. Samples were analyzed for N2O on a GC fitted with a PDID detector with He as the carrier gas. We compared zero-N with UAN injection based on nitrate soil test and with UAN injection based on spectral reflectance. Three chamber measurements were made in each 175-m long plot. Treatments were replicated three times except for zero-N. Denitrification data were not normal and were log-transformed for ANOVA. Fluxes ranged averaged 0.9 g N2O + N2-N/ha/hr. Between 15 June and 2 August there were significant treatment differences on 30 June only. On that date denitrification fluxes were 2.26, 0.13, and 0 g N2O + N2-N/ha/hr for soil test-based N injection, reflectance-based injection, and zero-N, respectively. However, the seasonal amounts of N lost through denitrification were lower than expected, and were not agronomically significant.