The objective of this study was to estimate the regional fluxes of the sensible and latent heat fluxes over tropical forest and pasture areas in Amazonia. These estimates have been derived from a horizontal homogenous energy and humidity budget equation (slab model) for the boundary layer and compared with fluxes measured with an eddy correlation equipment. The data-set used consist of 18 days of rawinsoundings made at 8:00, 11:00, 14:00 and 17:00 Local Time. during the Rondonia Boundary Layer Experiment (RBLE). This experiment has been carried out during the dry season (july 93 and august 94) at the Ji-Parana area (southwest of Amazonia in the Rondonia state). The potential and the specific humidity profiles used were representative of the different covered surface (tropical forest and pasture). The time series of the soundings (potential and specific humidity) and the growth of the boundary layer were introduced to the slab model and the associated surface fluxes were computed by its inverse form. The root mean square (RMSD) for the sensible heat flux was 39.5 W.m-2 for the forest site and 67.2 W.m-2 for the pasture. For the latent heat flux, the RMSD was computed as 64.8 and 137.7 for the forest and pasture areas, respectively. Although the model still shown an overestimation of the surface turbulent fluxes over the forest site, it was able to represent the both sensible and latent heat fluxes in a reasonable way. However, the same was not true for the pasture site, where the advection term has to be include to improve the comparisons between estimates and observed fluxes. From these results, there are evidences that an enhanced turbulence or a thermal secondary circulation can occur over the pasture area and the measured surface fluxes (both sensible and latent) were not representative of the energy exchange within the atmospheric boundary layer in the Amazon region