This paper addresses an integrated inventory and routing problem of two-echelon distribution system for a chain enterprise which consists of a central distribution center and a group of retail stores. Given daily demand forecasts for each store, distance, cost parameters, and various inventory and transportation constraints, the enterprise seeks to determine a short-term inventory of each member and the routing decisions among members of the enterprise, with the objective of minimizing the overall average inventory and transportation cost of the enterprise. Considering the change of total system cost resulting from the distribution plan establishment, a mathematical programming model of is proposed for the considered integrated inventory and routing problem, and a efficient two-phase heuristic algorithm solving this problem is also developed. The effect of the established model and the efficiency of its heuristic algorithm are illustrated by two set of computational experiments.