One of the most distinguishing features of environmental ethics has been the effort to develop a nonanthropocentric intrinsic value theory, that is, a definition of the good which is not dependent upon some quality particular to humanity, a definition of the good whereby properties found in the terrestrial, nonhuman world are constitutive of that definition. In this paper, I argue that major nonanthropocentric theories suffer from arbitrariness. I argue through the use of representative thinkers that much nonathropocentric theory has committed (...) the naturalistic fallacy because it has deployed various forms of empirical naturalism, and that to meet this challenge nonanthropocentrism must employ a form of metaphysically based nonanthropocentrism. I do not argue that the naturalistic fallacy is valid. Rather, I show that a sample of major thinkers, representative of a logically exhaustive set of possible evasions of the naturalistic fallacy, all fail to evade the fallacy. Further, I show that the failure of this set of possible evasionsleaves but one evasion possible, namely, ethical theory grounded in metaphysics. Finally, I recommend “process” metaphysics as the most promising metaphysical ground for environmental ethics, assuming the validity of the naturalistic fallacy. (shrink)