To monitor and model the evolution of urban landscapes, we develop a brightness–darkness–greenness (B–D–G) model. It is based on the vegetation–impervious surface–soil (V–I–S) model, proposed by Ridd (1995) to simplify urban environments to three basic ground components. The model integrates the knowledge of urban landscape composition and spectra of remote sensing. The B–D–G model is a fast and effective method to analyze urban landscape composition and its evolution based on remotely sensed images, by employing an explicit endmember evolution implication via the endmember spectrum dynamics. We verify this new method through in situ measurements of spectrum and high resolution images. Then, B–D–G model is used to detect the pattern and types of urban renewal. Despite some limitations, B–D–G model provides a new perspective of modeling urban dynamics and monitoring urban landscape evolution.