This paper deals with the problem of a firm that searches the optimal location of plants under the following conditions: (1) Transport costs for delivery are incurred by the firm; transport costs between a plant and a customer are linear. (2) The customers' demands are known and have to be met; customers are scattered. (3) The possible plant locations are known. (4) The production cost, at each possible plant, may be divided into two parts: (a) a fixed charge supported only if there is some production; (b) a linear unit cost. The rigorous treatment of the problem involves the production and transport costs' being minimized for each combination of plant locations. This paper presents an algorithm using the special features of the problem that reduces the number of combinations of plant locations to study. A heuristic method based on market simulation is also shown to work pretty well.