A newly developed method, which involves the use of satellite measurements of energy released by fires, was used to estimate smoke emissions in the United States (US) Southern Great Plains (SGP). This SGP region was chosen because extensive agricultural and planned burning occurs there annually. Moderate resolution imaging spectroradiometer (MODIS) aerosol optical depth (AOD) and fire radiative energy (FRE) release rates (R-FRF), acquired in 2004 from the Terra and Aqua satellites, were used to derive a FRE-based smoke emission coefficient (Ce kg MJ-1), which when multiplied by RFRE (MJ s-1) gives the rate of smoke emission (kg s-1). Correlations between the smoke emission rates and the RFRE were significant for Terra-MODIS (R2 = 0.645, n = 146, p < 0.0001) and Aqua-MODIS (R2 = 0.752, n = 178, p < 0.0001). Furthermore, the Ce values derived independently from Terra and Aqua were in close agreement, and the average Ce for this area is 0.049 ± 0.024 kg MJ-1. A Monte Carlo (MC) probabilistic approach was used to approximate uncertainties from the smoke emission and resulting Ce. For the first time, smoke emission estimates have been derived for the US SGP using observations of energy released by fires. Although more work is necessary, the present study demonstrates the feasibility of using RFRE for smoke emission estimation in that region. Burning peaked during the spring and fall seasons. Moreover, qualitative examination of smoke emission patterns side-by-side with local air-quality measurements indicated that the impact of smoke from local biomass burning activities was significant on the regional airquality. © 2008 Elsevier Ltd. All rights reserved.