Model of input-output balance (MIOB) is interconnected with model of territorial manufacturing locations, consumption and transportation of products in the following way. In MIOB total production and consumption volumes on the country are determined for each kind of product. Along with it, unit costs of transport services for its transportation are used, but it isn’t taken into account that their value depends on territorial location of places of producing and consuming products being transported. Moreover, not only total volume of transport activities in this field depends on these locations, but distribution of that volume by territorially independent or partially interchangeable element of the transportation system as well. In its turn, it determines the necessity to develop those of them that prove to be “bottlenecks”. Transport elements’ development requires investment in territorially fixed elements of the transportation system and these investments must be consistent with the total volume of investment in transport determined in MIOB. A system of models is constructed, in which MIOB and model of choosing territorial manufacturing locations are interconnected. From MIOB to model of territorial manufacturing locations total for the country production and consumption volumes, as well as investment volumes are transferred, and in the location problem these total values are distributed on the country’s territory so that total production and transportation costs were minimized. In the opposite direction values of average transportation distances determined by the chosen location are transferred, as well as a list of transport objects, capacity of which must be increased in order to realize the chosen location. It’s shown that the problem of choosing manufacturing and consuming locations represents a problem of linear programming of extremely large dimensions; however it has a special block structure allowing organizing an effective computing process. The equations which connect between themselves the separate blocks are divided into parts, corresponding to each block through introducing additional variables and their values are improved iteratively on the base of using dual estimates of blocks limitations. Verification and calibration of the model are foreseen through making calculations according to the data of a reporting period and comparing the data received as the result of calculation and the reporting data.