at the operational scale of forest planning, the location of roads, in a complex terrain, is a decision of major economic importance. the objective is to locate a network of roads connecting pre-selected cut-blocks, at minimal cost, subject to a host of feasibility constraints. the location of the pre-selected cut-blocks is a made at the tactical level of planning, and the decision is made with relative indifference to the cost of the connecting these blocks at the operational scale of planning. the objective of this thesis is to design and evaluate an optimization model by which the location of an operationally feasible road-network can be made simultaneously with the selection of cut-block locations. the underlying assumption of this undertaking is that the optimal locations of both cut-blocks and operational roads are interdependent our methods are summarized in three steps: first, an exhaustive library of candidate roads is generated, using an operational-scale, road location model; second, a subset of these roads is selected in a tactical planning model which optimally schedules the harvest of cut-blocks and location and construction of roads simultaneously; and third, the tactical solution is mapped, enabling inspection of road-locations at the scale of 50m x 50m. this innovative approach was tested on a portion of the kenogami forest, in ontario, and the results demonstrate its feasibility and ability to reduce road-construction costs. we conclude that the approach is useful for three reasons: first, it can reduce the total cost of constructing a road-network; second, it facilitates innovation in tactical planning by making operational road-location planning methods relevant to tactical level modeling; and finally, it can be practiced on large forests.