In this paper we analyze the relationship between capital and energy through cross price elasticities. First, we extend Thomsen's (2000) methodology in order to link the short and long run in a panel data setting, by including an equation for the motion of capital. Then, by using an expansive industry-level data set and two functional forms, we show clear evidence of long run complementarily in all the analyzed industries, and with respect to the different types of capital that we consider (buildings and machinery). We identify the industries with the greatest degree of dependence between energy and capital. These are therefore, the industries in which a policy of increasing energy prices via taxes to reduce energy consumption may have a serious effect, reducing their investment levels. Hence we recommend that a better governmental policy would be to encourage technological diffusion.