The title implies a clear separation of two approaches to the understanding of energy metabolism, which seldom, if ever, applies. Those empirical relationships which have stood the test of time were based on the concept of some uniformity of physiological function arising from an underlying and universal mechanism. Conversely the most sophisticated computer simulations of energy metabolism from cellular kinetics ultimately depend on empirical relationships which can be validated only by accurate measurements of whole-body energy expenditure.However, there are marked differences of philosophy and of purpose between empirical and mechanistic approaches. These were succinctly described by Baldwin & Miller (1989), and the author can do no better than to paraphrase two paragraphs of their comments.'The term empirical is appropriate to describe models/equations wherein data obtained in input:output animal experiments were used to parameterize equations. In a strict sense the only restriction imposed on defining empirical equations is that the equations should best describe relationships among two or more variables, i.e. the equations need not imply anything about underlying/metabolic relationships. Many models based on empirical relationships have been developed and have been useful when applied carefully and with appropriate recognition of the limitation that such equations only apply within the range and conditions of the collected data used to parameterize the model.Mechanistic models/equations, by definition are based on our understanding of underlying cause-and-effect relationships and should apply to a wide range of conditions. The use of the term mechanistic implies that our knowledge of the system is complete and, further, that sufficient biochemical, physiological and metabolic data are available to parameterize all the mechanistic equations in the model. This is rarely, if ever, the case. ' The main purpose of empirical relationships in the field of energy metabolism, as in other aspects of biological science, has been to identify and quantify the degree of uniformity of biological function within and between species. The secondary purpose has been to apply these relationships in order to determine energy requirements for different species in relation to specific physiological status, and this has been the cornerstone of development of feeding systems for animal production and of recommended dietary intakes for humans. Most notable among the empirical relationships are those relating to the allometry of growth and organ development, the surface area law and the inter-specific relationship between basal metabolism and metabolic body size. During the 1970s the increased availability of personal computers, coupled with the vast data pools which had been accumulated during the previous 20 years, gave an impetus to the development of complex factorial models such as the Edinburgh pig model (Whittemore available at https://www.cambridge.org/core/terms. https://doi