The main objective of this study includes to evaluate the SAR and optic image fusion performance for image classification in a complex urban environment. The Terra SAR-X (TS-X) SAR image and Quickbird optical data is used as the classifier inputs for land cover/use classification. Initially, multispectral imagery and high-resolution PAN imagery of Quickbird data were fused. After removing the SAR speckle from the SAR image it was fused by the pan sharpened optic image. Both pan sharpened optic image and fused SAR image were classified using the same training set in order to remove the uncertainty caused by training location. The urban environment selected for performance test of proposed approach involves buildings with various shapes and surface materials and some buildings may appear indistinguishable from roads and pavements. Due to this complexity in the region each building type is classified into different classes and then aggregated into one class. Each land cover/use classes are then evaluated by an error matrix for both classification result. The classification result of fused SAR image slightly changes from classification result of pan sharpened optic image for complex urban environment, and it performs well for all types of urban classes.