Isoprene emission in the eastern United States was estimated by a dry deposition and biogenic emission model. The algorithm for isoprene emission is based on the integration of leaf-level isoprene emission with respect to whole-canopy leaf area index. The unique feature of this algorithm is the use of satellite-observed reflectance to derive the leaf area index, which results in considerable detail in the spatial patterns. To evaluate model results, a comparison was made to observations at Harvard Forest. The model-estimated emission rates matched observations reasonably well. Further studies were conducted by simulation of ozone concentrations over eastern United States for an Ozone Transport Assessment Group (OTAG) case by using two methods of estimating isoprene emissions: a standard application of Biogenic Emissions Inventory System version 2 (BEIS2) and an application using remote sensing data from satellite. The Mesoscale Meteorological Model version 5 (MM5) was coupled with the dry deposition and biogenic emission model to generate emissions for July 7 and 8, 1995, days with significant ozone episodes. The satellite-derived isoprene emission derived by using meteorological data from MM5 tended to be higher than from results BEIS2 in the southwestern part of the domain but slightly lower along the East Coast. To examine the effect of uncertainties in isoprene emission on ozone formation in this region, three cases were examined. Case 1 was a standard OTAG case with the combination of BEIS2 isoprene emissions and OTAG meteorological data generated by Regional Atmospheric Modeling System (RAMS). Case 2 replaced BEIS2-derived isoprene emission with satellite-derived emission, with the intent of examining the effects of isoprene emission on ozone formation. Case 3 represented the current modeling approach of using the dry deposition and biogenic emission model together with MM5 simulations. The results for Case 2 showed that isoprene emissions affected ozone concentration in the northeastern United States, but had no noticeable impact in the southeastern United States except near large NOx point sources. A comparison of Case 3 and Case 1 indicated that the pattern of ozone concentration derived by using the current modeling approach of isoprene emission simulation and MM5-generated meteorological fields was very similar to the Case 1 OTAG standard results involving BEIS2 and RAMS meteorological data. The maximum value of ozone concentration was 114 ppb for Case 3 versus. 116 ppb for Case 1 on July 8, 1995. The meteorological fields, however, had very strong effects on spatial patterns of ozone concentrations. For example, the maximum value of ozone concentration was as large as 176 ppb when the MM5 simulations were made without four-dimension data assimilation, versus 116 ppb for the OTAG case on July 7, 1995.