Estimates of NOx (NO+NO2) emissions from temperate grasslands range from 0.003 to 101 ng NO-N m-2 s-1 (average 4.17 ng NO-N m-2 s-1). As a result of this uncertainty, the potential contribution of soil NOx emissions to the global budget and to nitrogen loss for this ecosystem is unclear. The few existing measurements are short-term observations of growing-season fluxes. We measured soil NOx emissions from six sites within the shortgrass steppe at the Central Plains Experimental Range in Colorado from June 1994 through October 1995. The soils at these sites provided a range of texture (from a sandy loam to a clay loam) and soil moisture. Mean NOx emissions over the sampling period ranged from 2.6 to 5.7 ng NO-N m-2 s-1 from the four unfertilized sites. Temperature was the dominant control on seasonal variations in NOx fluxes. Seasonal fluxes were highest in the summers (5.4 to 10.5 ng NO-N m-2 s-1) and lowest in the winter (0.2 to 1.5 ng NO-N m-2 s-1). The winter NOx emissions contribute up to 25% to the mean annual flux. Water-filled pore space (WFPS) alone was a poor predictor of NOx emissions; however, peak NOx emissions were found near the field capacities for these soils (32--35% WFPS for coarse soils and 66% WFPS for fine-textured soils). Water additions produced large (22--51 ng NO-N m-2 s-1) but short-lived (24 hour) pulses of NOx emissions that were independent of both the amount of water added and the number of antecedent dry days. Short-term increases in NOx flux stimulated by wetting are significant, and increase the summer estimate of NOx emissions 8 times estimates calculated from periodic sampling. Nitrogen applied in previous studies, 5 to 12 years earlier, increased the average annual NOx emissions approximately 1.5 times. Extrapolating our estimate to similar systems around the world, we estimate that grassland ecosystems, globally, emit 1.0 Tg N yr-1 as NOx. A current estimate of NOx emissions from grassland soils is 0.6 Tg N yr-1 [Davidson, 1991]. ¿ 1998 American Geophysical Union