We measured the time course of skeletal muscle actin polymerization at different actin concentrations. In 0.1 M KCl with 1 mM Mg2+, log/log plots of the rate of the early, slow phase of polymerization vs. actin concentration were linear with slopes from 1.0 to 1.3. Computer-assisted calculations of similar curves from theoretical models with different sizes for the nucleus showed that no simple model gave a log/log plot with a slope less than 1.5. Addition of a first-order, monomer activation step before nucleation allowed models of any reasonable nucleus size to have a slope of 1. This is the first evidence that such a step is part of the kinetic pathway for actin polymerization. In 0.1 M KCl with 0.2 mM Ca2+, log/log plots of the rate of the slow phase vs. actin concentration were linear with slopes from 2.0 to 2.5. Monomer activation was not necessary to account for this slope. However, fits of kinetic curves calculated from theoretical models to experimental kinetic curves showed that filament fragmentation was important to achieve a good fit, confirming the finding of Wegner and Savko [Wegner, A., & Savko, P. (1982) Biochemistry 21, 1909–1913]. Our fit procedure also allowed us to estimate the size of the nucleus and the rate constants for activation, nucleation, and fragmentation. In 0.1 M KCl with 1 mM Mg2+, the nucleus was a dimer or trimer, and nucleation was fast. In 0.1 M KCl with 2.0 mM Ca2+, the nucleus was a trimer, and nucleation was slow.