Abstract: In order to establish an eco-friendly energy structure and a comprehensive energy saving system and adapt to the long-term goal of sustainable development, the index decomposition model was utilized to calculate and analyze the influence of structure and efficiency on energy saving from 1996 to 2010 in Beijing. The energy consumption per unit of GDP was used as an important indicator of energy consumption and energy saving in Beijing. The vector autoregression (VAR) model was developed, through incorporating the impulse response analysis and variance decomposition, to analyze the dynamic relationship among the industrial structure (i.e., the amount of energy saving due to structural adjustment), efficiency (i.e., the amount of energy saving due to technological progress) and energy consumption per unit of GDP in Beijing. The VAR model was validated using the actual energy consumption per unit of GDP in Beijing from 2011 to 2013, and the error was found to be lower than 5%. The results showed that during 1996-2000, industrial structural adjustment played the most important role in energy saving, accounting for 55.60% of the total energy saving, mainly relying on the adjustment between the secondary and tertiary industries. In contrast, the energy saving resulting from technological progress was larger during 2001-2005 and 2006-2010, accounting for 81.35% and 85.41% during these two five-year plan periods, respectively. The variance contributions of the energy consumption per unit of GDP, industrial structural adjustment and technological progress to the energy consumption per unit of GDP were 13%, 39% and 48%, respectively. Moreover, both the industrial structural adjustment and the technological progress had a five-year periodic impact on energy saving, although this impact gradually weakened and vanished after a decade. The industrial structural adjustment mainly affected the energy consumption per unit of GDP during the first two and half years, in particular over the second year; however, technological progress played a dominant role in the energy consumption per unit of GDP from the third to fifth year, in particular over the third year, with a two-year delay. Thus, the evaluation of the effects of industrial structural adjustment and technological progress on energy saving should take into consideration the different impact of industrial structural adjustment and technological progress on energy consumption per unit of GDP.