In this article, a new technique of simulating and analyzing dynamic changes in landscape eco\|patterns is proposed based on grey\|situation decision\|based cellular automata. When only incomplete information is available, such a method can add credibility and feasibility to the dynamic modeling. Based on a case study on the middle reach of the Nujiang River in Yunnan province, three factors, including concentration of cellular neighbors, land suitability, and human disturbance, are found to be the major drivers of cellular transition, whose weight changes along with spatial and temporal alterations in accordance with the dynamic properties of such a complex system as a landscape. Randomness of transition in simulating is also taken into consideration by employment of the Monte Carlo method. Calculation results show that general tendencies seen in a simulated future scenario conform to the actual processes, not only simulating the micro self\|organizing mechanisms of landscape units, but also reflecting to some extent the macro social and economic impacts. Hence, it can be concluded that such an approach is more problem\|targeted, representative, and accurate than methods currently in use.