This thesis aims to develop a mathematical formulation to an integrated and flexible logistics network design that include choices of facility locations, transportation, production and inventories. The network designs consider vendors, plants, warehouses and transportation modes choices. The proposed model considers products assignment to plants and warehouses, inventory holding and procurement costs. The mathematical formulation of the model is a Mixer Integer Non Linear Program (MINLP) problem, referring to a single period with stochastic demand. The problem is NP-Hard and we used the Outer-Approximation algorithmic as the method to resolve the model proposed. We tested the algorithmic for three different instances (scenarios).