Energy system optimization models (ESOM) simulate energy and emissions changes under different economic and technological scenarios or prospective policy cases. ESOMs and larger integrated assessment models (IAMs) are increasingly being used to project future physical resource demands, but the integration of (non‐energy) physical resource flows or life cycle data into IAMs is far from complete. In this work we demonstrate a method to harness results from the National Energy Modeling System developed by Energy Information Administration (EIA), combined with imputed commodity prices from the UN COMTRADE database, in order to present detailed projections of the physical economy of the United States to 2050. Mass flow results for nine separate scenarios are presented, covering all extraction sectors and manufacturing sectors, with additional disaggregation possible to 4,601 commodities. Results are compared with previous estimates of physical resource flows through the US economy that utilized historical statistics or alternative modeling methods. Overall, the physical resource intensity of the US economy is projected to decrease by an average of 28% per unit of GDP by 2050, suggesting continued decoupling of physical resource use from economic output, but increase by an average of 25% on a per capita basis. These projections have implications for physical resource planning, particularly for materials that have constrained domestic supplies. We also investigate and discuss sources of potential bias and uncertainty in the imputed price estimates and suggest several opportunities to harness the physical resource flow projections for future resource modeling and industrial ecology research. This article met the requirements for a gold‐gold JIE data openness badge described at http://jie.click/badges.