We use an enclosure technique to monitor soil NO and N2O emissions during early summer regrowth of Bermuda grass (Cynodon dactylon) on sandy loam in a humid, subtropical region of southern Texas. The evolution of both gases was substantially higher from plots harvested at the beginning of the experiment and fertilized 5 days later with 52 kg N ha-1 as (NH4)2SO4 than from plots not harvested or fertilized. Emission of NO, but no N2O, was stimulated by clipping and removing the grass, probably because eliminating the shading provided by the dense grass canopy changed these plots from cooler to warmer than unharvested plots, thereby stimulating the activity of soil microorganisms responsible for NO production. Neither gas flux was significantly affected by application of N until the next rainfall dissolved and moved the surface-applied fertilizer into the soil. Immediately thereafter, emissions of NO and N2O increased dramatically to peaks of 160 and 12 g N ha-1 d-1 , respectively, and then declined at rates that closely paralleled the nitrification rate of added NH4+, indicating that the gases resulted from the activity of nitrifying microorganisms, rather than denitrifiers. Nitric oxide emissions during the 9-week measurement period averaged 7.2 times greater than N2O emissions and accounted for 3.2% of the added N. The data indicate that humid, subtropical grasslands, which not only have large geographical extent but also have been subject to intense anthropogenic disturbance, contribute significantly to the global atmospheric NOx budget. ¿ American Geophysical Union 1992