-We review metabolizable energy (ME) concepts and present evidence suggesting that the form of ME used for analyses of avian energetics can affect interpretation of results. Apparent ME (AME) is the most widely used measure of food energy available to birds. True ME (TME) differs from AME in recognizing fecal and urinary energy of nonfood origin as metabolized energy. Only AME values obtained from test birds fed at maintenance levels should be used for energy analyses. A practical assay for TME has shown that TME estimates are less sensitive than AME to variation in food intake. The TME assay may be particularly useful in studies of natural foods that are difficult to obtain in quantities large enough to supply test birds with maintenance requirements. Energy budgets calculated from should be expressed as kJ of AME and converted to food requirements with estimates of metabolizability given in kJ AME/g. Energy budgets calculated from multiples of basal metabolic rate (a component of maintenance energy), however, should be expressed as kJ of either TME or net energy depending on ambient temperature. Energy units should be stated explicitly to improve comparability and in some cases accuracy of energy analyses. Metabolizable energy (ME; see Appendix for explanation of this and other abbreviations) is a measure of the energy available to birds from their diet (Vohra 1966). In avian energetics, ME is used to convert daily energy budgets (DEB; see, e.g., King 1974, Kendeigh et al. 1977) to weight of food required to supply energy needed by individuals or populations (Kendeigh et al. 1977, Wiens and Dyer 1977). ME has been used in this way for DEBs based on existence metabolism (EM; Kendeigh et al. 1977) and multiples of basal metabolic rate (BMR; King 1973, 1974, Ricklefs 1974, Prince 1979). ME can be expressed as either apparent (AME) or true (TME) metabolizable energy (Harris 1966). The form of ME used in modeling calculations has almost always been AME, although not often specified as such (Ricklefs 1974, Wiens and Dyer 1975, 1977, Craig et al. 1979, Prince 1979, Raveling 1979, Ashkenazie and Safriel 1979, Sugden et al. 1981). AME has been the traditional measure of ME in studies of birds because, until recently, standard feeding trials did not separate total excretory energy into independent estimates of fecal plus urinary energy (FE + UE) and of metabolic fecal plus endogenous urinary energy (fecal and urinary energy of nonfood origin; FErn + UE,; Sibbald 1976). However, TME correctly recognizes FErn and UE, as metabolized energy (Harris 1966, Owen and Reinecke 1979) and is, therefore, a more direct measure of energy availability. Our purposes here are to provide a brief review of ME concepts and to show that the form of ME used for energetics analyses can affect interpretation of results. AME is usually estimated with data obtained from test birds confined in cages. The birds are fed experimental diets over a number of days during which total food ingested and excreta voided are measured. This is the total collection method (Vohra 1972). Bomb calorimetry is used to determine the energy content of food and excreta, and AME (in kJ/g) is calculated from the equation: AME = [(GE,)(Qi) (GEe)(Qe)]/Qi (1) where GE, and GE, = the gross energy density of food that is eaten (intake) and excreta (in kJ/g), and Qj and Q, = the quantity of food eaten and excreta produced (in g). Alternatively, a nondigestible chemical (Vohra 1972) or crude fiber (Almquist and Halloran 1971) can be used as a tracer to relate excreta production to food intake: AME = GE, (% T,/% Te)GE (2) where % T, and % T, = percent tracer in the food and excreta. Expressions of energy availability that use digestive efficiency (Ashkenazie and Safriel 1979, Raveling 1979), assimilation eff ciency (Ricklefs 1974), or metabolizable en-