Actobindin, a 25,000-dalton dimeric protein purified from Acanthamoeba castellanii was previously shown to form a 1:1 molar complex with both Acanthamoeba and rabbit muscle G-actin with KD values of about 5 muM and 7 muM, respectively, and not to interact with F-actin (Lambooy, P. K., and Korn, E. D. (1986) J. Biol. Chem. 261, 17150-17155). We now find that actobindin is a much more potent inhibitor of the early phases of polymerization of both Acanthamoeba and muscle G-actin than can be accounted for by its binding to G-actin. Actobindin inhibits the polymerization of both G-ATP-actin and G-ADP-actin, and has little, if any, effect on the rate of ATP hydrolysis that accompanies polymerization of G-ATP- actin. The kinetics of actin polymerization in the presence of actobindin are qualitatively consistent with the postulation that actobindin binds reversibly to and inhibits the elongation of an intermediate between G-actin and F-actin, perhaps a small oligomer(s), or a species in equilibrium with such an intermediate. This hypothesis implies the, at least transient, existence of an actin species with properties different than those of monomers and filaments. Actobindin may, then, provide a useful experimental tool for investigating the still relatively obscure early steps in actin polymerization. Irrespective of its mechanism of action, actobindin might serve in situ to reduce the rate of actin polymerization de novo while having relatively little effect on the rates of elongation of existing filaments or from actobindin- resistant nucleating sites.