Although Acanthamoeba actobindin binds actin monomers, its inhibition of actin polymerization differs from that of a simple monomer-sequestering protein in that actobindin inhibits nucleation very much more than elongation [Lambooy, P. K., & Korn, E. D. (1988) J. Biol. Chem. 263, 12836-12843] and can induce the accumulation of actin dimers in stoichiometric excess of the actobindin concentration [Bubb, M. R., Knutson, J. R., Porter, D. K., & Korn, E. D. (1994) J. Biol. Chem. 269, 25592-25597]. We now describe a "catalytic" model for the interaction of actobindin with actin monomer that quantitatively accounts for the effects of actobindin on the kinetics of actin polymerization de novo and the elongation of actin filaments. We propose that, in a polymerizing buffer, actobindin binds to two actin subunits forming an heterotrimeric complex that is incompetent for nucleation, self-association, and elongation. Actobindin can, however, dissociate from this complex, leaving a novel actin dimer that can participate in elongation but remains incompetent for nucleation and self-association. Under appropriate conditions, the concentration of this novel actin dimer can exceed the actobindin concentration; thus, the model is catalytic rather than stoichiometric. The experimentally observed time course of actin polymerization de novo, the rate of elongation of filaments, and the amount of actin dimer formed as a function of actobindin concentration are all consistent with the catalytic model and inconsistent with the stoichiometric model. The rate of actobindin-induced actin dimer formation is consistent with the hypothesis that the rate-limiting step is this pathway is the formation of a precursor heterotrimeric complex.