In order to scientifically predict the economic indicators that enterprises are concerned about, so as to make correct decision-making plans for the future behavior of enterprises, it is necessary to use appropriate mathematical models and methods to quantitatively study the economic activities of enterprises. Based on the complex, changeable and many random factors of economic activities, the corresponding Markov chain model is established for two common economic problems, and the relevant theory of Markov chain is applied to skillfully construct the transfer probability matrix, and the problem can be solved quickly through simple matrix operations. Examples show that the Markov chain model and method are feasible and practical in the analysis of enterprise economic activities, and can be widely used to solve common forecasting and decision-making problems in enterprises.