The iron and steel industry generates around 10 % of global greenhouse gas emissions. The bulk of the emissions originates from the iron ore reduction. In this reduction, coal is used as a reagent. Steelmakers could switch to hydrogen-based direct reduction using hydrogen instead of coal as a reagent to reduce iron ore to pig iron. This would eliminate the CO2 emissions from the equivalent process in a traditional blast furnace. However, the process requires massive amounts of electricity. This paper looks at the economics of such a switch to “green steel.” We assess a marginal increase in the production of a hypothetical green steelmaker. We also undertake an investment appraisal of a green plant, based on an ongoing installation in Northern Sweden, but also briefly consider a possible/planned investment in the US. This appraisal is complemented by computing the survival function for the net present value in a systematic sensitivity analysis. It seems highly unlikely that a green steel plant can be socially profitable. If the green plant displaces conventional steel produced within the European Union’s cap-and-trade system for greenhouse gases, total emissions remain more or less unaffected; permits and emissions are simply reshuffled. Hence, if end-users of green steel pay a premium, they might pay for an illusion.