Multi-angle remote sensing has been proved useful for mapping vegetation community types in desert regions. Based on Multi-angle Imaging Spectro-Radiometer (MISR) multi-angular images, this study compares roles played by Bidirectional Reflectance Distribution Function (BRDF) model parameters with those played by topographic parameters in improving vegetation community type classification at the Jornada Experimental Range and the Sevilleta National Wildlife Refuge in New Mexico, U.S.A. The BRDF models used are the Rahman-Pinty-Verstraete model and the RossThin-LiSparseReciporal model. MISR nadir multi-spectral reflectance is considered as baseline because nadir observation is the most basic remote sensing observation. The BRDF model parameters and the topographic parameters are considered as additional data. The BRDF model parameters are obtained via inversion of the Rahman-Pinty-Verstraete model and RossThin-LiSparseReciporal model against the MISR multi-angle reflectance. The topographic parameters are derived from Digital Elevation Models of the Jornada and the Sevilleta. In order to confirm whether increased accuracy is from additional data, two classifiers, Maximum Likelihood Classification algorithm and Support Vector Machine algorithm, are used in this research. In total, 32 classification experiments were carried out by the two classifiers on 16 datasets, which were formed by various combinations of the MISR nadir multi-spectral reflectance, the BRDF model parameters and the topographic parameters. These experiments show the following findings: 1) both the topologic parameters and the BRDF parameters can provide meaningful additional information for this semi-arid vegetation mapping; and 2) in the case of a semi-arid environment, the BRDF parameters are slightly more efficient than the topologic parameters. This study suggests that the MISR BRDF model parameter products have great potential to be used as additional information for vegetation mapping.