A global three-dimensional (3-D) chemistry and transport model was used to demonstrate that a factor of 2 decrease in isoprene and monoterpene emissions results in significant (10--30%) changes in predicted concentration distributions of compounds such as OH, MPAN, NOx, H2O2, O3, and CO. Isoprene and monoterpenes were predicted to have a particularly strong impact on tropical regions, including central Africa. The 1996 Experiment for Regional Sources and Sinks of Oxidants (EXPRESSO) study included a number of experiments that improved our ability to predict isoprene and monoterpene emissions from central Africa. The results of these experiments have been incorporated into an isoprene emission model that predicts hourly emissions on a spatial scale of about 1 km2. The model uses procedures that are suitable for estimating global emissions but uses regional measurements to accurately parameterize the model. Comparisons with above-canopy aircraft and tower flux measurements demonstrate that the model can estimate emissions within a factor of 2 for regions where ground measurements of model inputs are available. The annual central African isoprene emission predicted by our revised model (35 Tg C) is only 14% less than that predicted by our earlier model, but there are considerable differences in estimates of individual model variables. The models differ by more than a factor of 5 for specific times and locations, which indicates that there are large uncertainties in emission estimates for at least some locations and seasons. The good agreement obtained for the EXPRESSO study field sites, however, suggests that the model can predict reasonable estimates if representative field measurements are used to parameterize the model. ¿ 1999 American Geophysical Union