"Introduction Concern over potential global warming@ due in part to the combustion of fossil fuels@ has stimulated interest in policies designed to reduce the emissions of greenhouse gases (GHG)@ principally carbon dioxide. A wide range of control instruments has been proposed. These include various demand management approaches aimed to encourage conservation (e.g.@ car pool requirements to lower total vehicle miles traveled)@ priceinduced incentives (e.g.@ carbon taxes and marketable permits)@ command and control (e.g.@ appliance efficiency standards) or end-of-pipe abatement (e.g.@ capture and storage of carbon dioxide).3 Price-induced mechanisms@ particularly carbon taxes@ presently receive wide attention by economists and policy makers as a possible GHG mitigation mechanism. This will likely continue into the future. Voluntary conservation efforts alone are deemed by some as having insufficient emissions reduction potential. As well@ many economists believe that price-induced incentives have lower marginal control costs than command and control or end-of-pipe abatement approaches over a wide range of emission levels. Finally@ many current economic models used to assess CO2 reduction impacts more easily accommodate price mechanisms than other control options. The first part of this paper undertakes an input-output (I/O) analysis of the impact of a carbon tax on the production cost of a full array of industries in the agricultural@ mining@ manufacturing@ transportation@ utility@ and service sectors of the u.s. economy. Since a carbon tax would likely induce substitution away from carbon intensive fossil fuels and I/O analysis does not allow for substitution among factors or prodQcts@ the role of I/O analysis in arialyzing its effects is constrained. However@ I/O analysis can reveal the initial cost impacts of a carbon tax on a detailed industry basis@ i.e.@ it can identify which industries would be put under the greatest pressure to undertake changes in their operations@ e.g.@ new capital expenditures or changes in the factor/product mix@ to mitigate the ""short-run"" increases in production costs. Looked at in another way@ the I/O estimates should be viewed as short-run estimates of total cost increases. A study by Charles River Associates and DRIlMcGraw-Hill (1994)@ however@ estimates a capital stock turnover rate of between 10 and 30 years for energy-intensive capital equipment. The long time period is due to the difficulty of retrofitting older equipment@ and the expense of prematurely retiring existing equipment. Hence@ the short-run may not be so short for some energy intensive industries. An implication is that the I/O cost estimates for some industries may be reasonably accurate over a significant period of time. The second part of this paper assesses the existing fuel switching capability in the manufacturing and electric utility sectors@ and what this implies for the interpretation of the I/O production cost estimates over the longer term. It also examines the potential trade impacts of a carbon tax on various energy-intensive U.S. manufacturing and natural resource industries. Such a comprehensive identification of the potential industry impacts of a carbon tax should provide useful information in the assessment of policy options to address potential global warming. 3Other strategies also exist such as expanding the carbon sink potential through reforestation and engineering climate change. Reforestation is thought by some to offer substantial potential for offsetting carbon emissions at relatively low cost."