In the Amazon Basin, biogenic, biomass burning, and urban emissions react photochemically to produce ozone, which is then transported by convection and regional circulations. As the Amazon basin becomes increasingly populated and developed for agriculture, the ability to predict regional ozone distributions will be critical to mitigating pollution that harms human health, agriculture, and natural ecosystems. However, major uncertainties remain in characterizing emissions, photochemical processes and pollutant transport in the Amazon basin. During the BARCA (Balanço Atmosférico Regional de Carbono na Amazônia) campaigns during the dry-to-wet (November and December 2008) and wet-to-dry (May 2009) transition seasons, basin-wide CO and O3 aircraft measurements were taken. In this study, these in-situ data, as well as remote sensing observations of O3, NO2, and CO, are used to evaluate the WRF-Chem and CCATT-BRAMS coupled chemistry and meteorology models, constrain emissions and analyze ozone budgets in the Amazon. The WRF-Chem (Weather Research and Forecasting with Chemistry) model developed at NOAA/ESRL was adapted in order to more realistically represent tropical meteorological and chemical processes. New WRF-Chem implementations include updated land cover and surface properties, soil moisture initialization from an off-line hydrological model, EDGAR4 urban emissions updated with vehicle emissions inventories for South American cities, and chemistry initial and boundary conditions from the MOCAGE global model. CCATT-BRAMS (Coupled Chemistry-Aerosol-Tracer Transport model coupled to the Brazilian developments on the Regional Atmospheric Modeling System) is currently in operational use for regional air quality forecasting at Brazil's National Institute for Space Research (INPE). Simulations were conducted with both models using RACM (Regional Atmospheric Chemistry Mechanism) and parameterizations for shallow and deep convection and 1D plume rise, with a 140/35km 2-way nested-grid configuration covering the Amazon Basin. The performance of the meteorological models was evaluated using flux tower and surface meteorological station data and the TRMM precipitation product. Simulated CO and O3 mixing ratios are compared with in-situ observations from the BARCA campaigns, OMI/MLS O3, OMINO2, and MOPITT CO products, and the model performance in reproducing both individual flight tracks and regional distributions is evaluated.