The purpose of this study was to assess feedstock logistics for a mobile pyrolysis system and to quantify the amount of soil loss caused by harvesting agricultural feedstocks for bioenergy production. The analysis of feedstock logistics was conducted using ArcGIS with the Network Analyst extension and model builder. A square grid methodology was used to determine biomass availability of corn stover and bioenergy sorghum in Texas. The SWAT model was used to quantify soil erosion losses in surface runoff caused by sorghum residue removal for bioenergy production in the Oso Creek Watershed in Nueces County. The model simulated the removal of 25, 50, 75, and 100 percent residue removal. The WEPS model was used to quantify wind erosion soil loss caused by corn stover removal in Dallam County. Nine simulations were run estimating soil loss for corn stover removal rates of 0 percent to 50 percent. The results of the SWAT and WEPS analyses were compared to the NRCS tolerable soil loss limit of 5 tons/acre/year for both study areas. The GIS analysis determined the optimum route distances between mobile unit sites were 2.07 to 58.02 km for corn and 1.95 to 60.36 km for sorghum. The optimum routes from the mobile pyrolysis sites and the closest refineries were 49.50 to 187.18 km for corn and 7.00 to 220.11 km for sorghum. These results were used as input to a separate bioenergy economic model. The SWAT analysis found that maximum soil loss (1.24 tons/acre) occurred during the final year of the simulation where 100 percent of the sorghum residue was removed. The WEPS analysis determined that at 30 percent removal the amount of soil loss starts to increase exponentially with increasing residue removal and exceeds the tolerable soil loss limit. Limited harvesting of biomass for bioenergy production will be required to protect crop and soil productivity ensuring a sustainable biomass source.