The present invention discloses a user demand response assessment method in an intelligent power grid environment. The method comprises the steps of: firstly, studying an interaction mechanism of a communication system in an intelligent power grid, establishing a structural model of an information bidirectional interaction system under an intelligent power grid framework, and determining a signal interaction manner; next, studying a long-term stable state demand response characteristic, and establishing a demand response assessment model based on a Markov chain; then, studying a short-term dynamic demand response characteristic, and by using a conditional probability distribution theory, establishing a demand response assessment model based on a Markov chain; and finally, performing an assessment on to-be-assessed objects in long-term and short-term demand responses. The method lays the foundation of further study on optimization and control of intelligent power grids, such as random optimization of power scheduling and electricity pricing.