This study intended (1) to develop a quantitative method to describe vegetation succession or change by remote sensing approach, and (2) to test the capability in clarifying the stability or instability of natural forests in landscape scales. With special attention on the coniferous forest, the attempt was conducted on the natural reserve of Changbai mountain in northeast China by using multi-temporal TM images. Two simple parameters were derived to represent succession rate or change extent. One was the mean change of radiance, and another was the number of pixels with changed reflection property. As conclusion, TM imagery is effective for detecting vegetation changes. The results also demonstrated that the so called climax is not a pure stand, which is generally assumed to be exclusively dominated by climax species, but rather a complex of shifting mosaic. The pioneer patches are permanent units in the forest community. This method is considered applicable for assessing the developing status or behavior of plant communities in large scales, like a life zone, under disturbance of global warming, especially for long-time span.