In this paper, we invert a canopy reflectance model using multispectral satellite data in order to estimate remotely the parameters of crown size, tree count density, cover, foliage biomass, and total standing biomass for nine coniferous forest stands on a transect of western and central Oregon (USA). Issues involved in the inversion of the model are topographic correction, component signature estimation, spatial pattern analysis, and direct biomass estimation. Retrieved parameters correlate well with observed values from ground measurements of the test sites. The model inversion technique is a useful tool for mapping and timber inventory of forests at large scales.