In this paper, we address a location-inventory-routing model for perishable products. The model deter-mines the number and location of required warehouses, the inventory level at each retailer, and the routestraveled by each vehicle. The proposed model adds location decisions to a recently published inventoryrouting problem in order to make it more practical, thus supporting the prevalent claim that integrationof strategic, tactical and operational level decisions produces better results for supply chains. Given thatthe model developed here is NP-hard, with no algorithm capable of finding its solution in polynomialtime, we develop a Genetic Algorithm approach to solve the problem efficiently. This approach achieveshigh quality near-optimal solutions in reasonable time. Furthermore, the unique structure of the prob-lem requires developing a new chromosome representation, as well as local search heuristics. Finally, ananalysis is carried out to verify the effectiveness of the algorithm.