The Swedish carbon dioxide tax was introduced in 1991, by adjusting the existing energy taxation to consider the carbon load of fuels. The tax was initially set at a general level of US $133* per ton carbon (tc). It was differentiated in 1993, with the result that industry paid US $43/tc while non-industrial consumers paid US $160/tc. This paper presents an ex post evaluation of the tax, using the criteria developed by OECD in 1997. The period under consideration is from 1990 to 1995. The effects of the tax must be seen in relation to other policy measures introduced. The effects of the tax vary across sectors. Biofuel use in the district heating sector increased from 36.7 PJ to 73.4 PJ, replacing primarily coal, thus leading to great carbon dioxide savings. Dynamic effects of the tax include development of new industry for refined wood fuels and extraction machinery. Transport has not been affected. Industry pays lower taxes on fossil fuels with the differentiated tax than it did before the tax was introduced, leading to increased fossil fuel use. While potentially cost effective measures in industry have thus been lost, international competitiveness has not been affected. The effect of the carbon dioxide tax on emissions depends on system assumptions; estimated abatement ranges from 0.5 to 1.5 million tons CO2 on a yearly basis.