Forest roads are known as the most important factor of sediment production in forest ecosystems. Because of the importance of this issue, several models developed based on empirical relationships to estimate sediment yield. In this study, the amount of annual sediment yield derived from a forest road network was estimated as tons per year using CULSED model in ArcGIS. Base spatial layers such as: topography, stream, roads, and culverts were used in this model. Additional data such as the bed rock, vegetation, average precipitation, road age, the amount of traffic, road surface and road gradient were used in the model. First, digital databases were developed and the model was run. Then places with high rates of sediment yield were specified and some culverts were designed in appropriate places to decrease the amount of sediment production. The results showed that the amount of sediment with regards to existing 38 culverts was 13.2 tons per year. Adding 16 proposed culverts in the areas with high sediment production resulted in 36% decrease in sediment yield. So, given the importance of reducing sedimentation from roads in the forest, this tool can be used before road construction to find optimal number and location of culverts by engineers. However, due to lack of available data and a native model, some uncertainty will be encountered that should be considered.