We summarize the recent information on field metabolic rates (FMR) of wild terrestrial vertebrates as determined by the doubly labeled water technique. Allometric (scaling) relationships are calculated for mammals (79 species), reptiles (55 species), and birds (95 species) and for various taxonomic, dietary, and habitat groups within these categories. Exponential equations based on body mass are offered for predicting rates of daily energy expenditure and daily food requirements of free-ranging mammals, reptiles, and birds. Significant scaling differences between various taxa, dietary, and habitat groups (detected by analysis of covariance with P < or = 0.05) include the following: (a) The allometric slope for reptiles (0.889) is greater than that for mammals (0.734), which is greater than that for birds (0.681); (b) the slope for eutherian mammals (0.772) is greater than that for marsupial mammals (0.590); (c) among families of birds, slopes do not differ but elevations (intercepts) do, with passerine and procellariid birds having relatively high FMRs and gallinaceous birds having low FMRs; (d) Scleroglossan lizards have a higher slope (0.949) than do Iguanian lizards (0.793); (e) desert mammals have a higher slope (0.785) than do nondesert mammals; (f) marine birds have relatively high FMRs and desert birds have low FMRs; and (g) carnivorous mammals have a relatively high slope and carnivorous, insectivorous, and nectarivorous birds have relatively higher FMRs than do omnivores and granivores. The difference detected between passerine and nonpasserine birds reported in earlier reviews is not evident in the larger data set analyzed here. When the results are adjusted for phylogenetic effects using independent contrasts analysis, the difference between allometric slopes for marsupials and eutherians is no longer significant and the slope difference between Scleroglossan and Iguanian lizards disappears as well, but other taxonomic differences remain significant. Possible causes of the unexplained variations in FMR that could improve our currently inaccurate FMR prediction capabilities should be evaluated, including many important groups of terrestrial vertebrates that remain under- or unstudied and such factors as reproductive, thermoregulatory, social, and predator-avoidance behavior.