Metabolic research has, like most areas of research in the life sciences, been affected dramatically by the application of transgenic technologies. Within the specific area of bioenergetics it has been thought that transgenic approaches in mice would provide definitive proof for some longstanding metabolic theories and assumptions. Here we review a number of transgenic approaches that have been used in mice to address theories of mitochondrial efficiency. The focus is largely on genes that affect the coupling of energy substrate oxidation to ATP synthesis, and thus, mice in which the uncoupling protein (Ucp) genes are modified are discussed extensively. Transgenic approaches have indeed provided proof-of-concept in some instances, but in many other instances they have yielded results that are in contrast to initial hypotheses. Many studies have also shown that genetic background can affect phenotypic outcomes, and that the upregulated expression of genes that are related to the modified gene often complicates the interpretation of findings.