Changes in material use, energy use and emissions profiles of industry are the result of complex interrelationships among a multitude of technological and economic drivers. To better understand and guide such changes requires that attention is paid to the time-varying consequences that technology and economic influences have on an industry's choice of inputs and its associated (desired and undesired) outputs. This paper lays out an approach to improving our understanding of the dynamics of large industrial systems. The approach combines engineering and econometric analysis with a detailed representation of an industry's capital stock structure. A transparent dynamic computer modeling approach is chosen to integrate information from these analyses in ways that foster participation of stakeholders from industry and government agencies in all stages of the modeling process-from problem definition and determination of system boundaries to generation of scenarios and interpretation of results. Three case studies of industrial energy use in the USA are presented-one each for the iron and steel, pulp and paper, and ethylene industry. Dynamic models of these industries are described and then used to investigate alternative carbon emissions and investment-led policies. A comparison of results clearly points towards two key issues: the need for industry specific policy approaches in order to effectively influence industrial energy use, fuel mix and carbon emissions, and the need for longer time horizons than have typically been chosen for the analysis of industrial responses to climate change policies.