Forest roads are an important part of forest management, both in terms of cost and impact on surrounding ecosystems. Existing tools to simulate the construction of forest roads have been designed for tactical or operational planning purposes, for relatively small areas (<10 000 ha) and small-scale topographic information. Hence, no forest road simulation tool properly exists to assist forest landscape ecology and management research. Here, we present the Forest Roads Simulation (FRS) extension for the LANDIS-II model—a spatially explicit landscape simulation model of forest succession and disturbances. The FRS extension simulates forest road networks via a least-cost path algorithm accounting for landscape structure, decision inputs, and forest road types. We demonstrate the accuracy with which the FRS extension reproduces several key characteristics of existing road networks in two managed regions in Quebec, Canada: road density, road position, and fragmentation of the landscape. The FRS extension is easy to parameterize, proposing many options for researchers to simulate forest road networks at a strategic level in managed landscapes. It can tackle new research questions investigating the effects of forest roads within management strategies, such as the cost of road construction and habitat fragmentation, across large management units and long planning horizons.