Picrate (trinitrophenol) is a unique uncoupler of oxidative phosphorylation. Unlike the commonly used uncouplers (e.g., 2,4-dinitrophenol, pentachlorophenol, m-chlorocarbonylcyanide phenylhydrazone, and 5-chloro-3-t-butyl-2'-chloro-4'-nitrosalicylanilide), picrate seems to penetrate the mitochondrial inner membrane very slowly. Consequently, it is ineffective when added to intact mitochondria or to mitochondria depleted of their outer membranes. In contrast, when added to phosphorylating submitochondrial particles prepared by sonication in which the inner membrane orientation is inside-out, picrate binds to the uncoupler-binding sites and uncouples oxidative phosphorylation. This unique property of picrate has made it possible to compare the potencies of picrate and dinitrophenol for (a) uncoupling and (b) increasing the proton permeability of submitochondrial particle vesicles. At 50% uncoupling concentration, dinitrophenol increased the proton permeability of submitochondrial particle vesicles by 9- to 12-fold. In contrast, at 100% uncoupling concentrations or higher, picrate augmented the proton permeability of the particles by only about 3-fold. These results indicate that facilitation of transmembrane proton equilibration does not determine the degree of uncoupling, and lead to the corollary conclusion that the magnitude of transmembrane proton gradient need not be the quantitative driving force for ATP synthesis.