Bursts of actin polymerization in vivo involve the transient appearance of free barbed ends. To determine how rapidly barbed ends might appear and how long they might remain free in vivo, we studied the kinetics of capping protein, the major barbed end capper, binding to barbed ends in vitro. First, the off-rate constant for capping protein leaving a barbed end is slow, predicting a half-life for a capped barbed end of approximately 30 min. This half-life implies that cells cannot wait for capping protein to spontaneously dissociate from capped barbed ends in order to create free barbed ends. However, we find that phosphatidylinositol 4,5-bisphosphate (PIP2) and phosphatidylinositol 4-mono-phosphate (PIP) cause rapid and efficient dissociation of capping protein from capped filaments. PIP2 is a strong candidate for a second messenger regulating actin polymerization; therefore, the ability of PIP2 to remove capping protein from barbed ends is a potential mechanism for stimulating actin polymerization in vivo. Second, the on-rate constant for capping protein binding to free barbed ends predicts that actin filaments could grow to the length of filaments observed in vivo during one lifetime. Third, capping protein beta-subunit isoforms did not differ in their actin binding properties, even in tests with different actin isoforms. A major hypothesis for why capping protein beta-subunit isoforms exist is thereby excluded. Fourth, the proposed capping protein regulators, Hsc70 and S100, had no effect on capping protein binding to actin in vitro.