It has been suggested that potentials measured with conventional microelectrodes in chemically or mechanically skinned muscle fibers arise from a Donnan equilibrium due to myofilament fixed charges. This hypothesis was tested in mechanically skinned frog (Rana pipiens) semitendinosus fibers by measuring the distribution potential (Ed) between fiber and bath with 3 M KCl-filled microelectrodes and the K+ activity gradient (aik/aok) with K+ ion-selective microelectrodes (KISE). If skinned fibers are a Donnan system, Ed should become more positive as pH is decreased, altering the fixed charge on the myofilaments. Consistent with this expectation, Ed was -4.4, -0.6, and +4.8 mV in ATP-containing solutions and -6.5, -2.2, and +8.4 mV in ATP-free solutions at pH 7, 6, and 5, respectively. Donnan equilibrium also requires that all mobile ionic species be in electrochemical equilibrium. In ATP-containing solutions, this was true for K+ at pH 7. At pH 5, however, KISE indicated that K+ was not in equilibrium; average Ed was 5.9 mV positive to the K+ equilibrium potential, and aik/aok was 1.04, while the Donnan prediction was 0.83. In contrast, KISE measurements in ATP-free solutions indicated that K+ was in equilibrium at all pH studied. Skinned fibers in ATP-containing media are not equilibrium systems because ATPase reactions occur. Under our conditions, frog myofibrils hydrolyze 0.4 and 0.08 mumol ATP/min X mg myofibrillar protein at pH 7 and 5, respectively. It is suggested that in the presence of ATP, Ed is a superposition of Donnan and diffusion potentials, the latter arising from differences in the mobilities of anionic substrate and products that diffuse through the charged myofilament lattice. A coupling to diffusion of K+, the predominant counter ion, is required for macroscopic electroneutrality. This coupling may be the origin of the nonequilibrium K+ distribution.