Concern has recently intensified regarding increases in the consumption of energy services that often follow energy efficiency improvements, a phenomenon widely called the 'rebound effect'. However, while some economists have precisely defined this as a metric, much discussion in academic and policy literature is imprecise, leading to confusion and miscommunication. This is especially so regarding direct 'rebound effects' in thermal retrofits of existing homes. This study surveys common usages of the term 'rebound effect' in domestic heating, identifying three main metrics, which employ different mathematical forms and therefore give different results, but are often lumped together. It defines these as the 'classic' rebound effect; the 'energy savings deficit', and the 'energy performance gap'. It then applies these to an empirical case study of three recently retrofitted 30-apartment buildings in Germany. It finds that each metric gives different results for identical situations, ranging from 2.0% to 29.9% for one building, 15.7% to 56.8% for the second, and 43.7% to 272.9% for the third. This may be one reason so-called 'rebound effect' results from various studies are so disparate. Nevertheless, specific uses are identified for each of the three metrics, provided their precise definitions are made clear. (C) 2013 Elsevier B.V. All rights reserved.