Single actin filaments stabilized with fluorescent phalloidin exhibit ATP-dependent movement on myosin filaments fixed to a surface. At pH 7.4 and 240C, the rates of movement average 3-4 ,um/s with skeletal muscle myosin and 1-2 jim/s with Dictyostelium myosin. These rates are very similar to those measured in our previous myosin movement assays. The rates of movement are relatively independent of the type of actin used. The filament velocity shows a broad pH optimum between 7.0 and 9.0, and the concentration of ATP required for half-maximal velocity is 50 ,AM. Evidence was obtained to suggest that movement of actin over myosin requires at most the number of heads in a single thick filament. This system provides a practical, quantitative myosin-movement assay with purified proteins.The lack of a simple in vitro assay for myosin movement has hampered progress in elucidating the mechanism by which actin and myosin couple ATP hydrolysis to mechanical work. We have sought to develop a totally purified system that would permit quantitative determination ofthe rate of myosin movement along actin.Several systems for observing movement of purified actin and myosin have been reported (1-6). The first quantitative measurement of rates of movement of purified myosin along actin in vitro were made by using the Nitella-based movement assay of Sheetz and Spudich (7). In that assay, myosin filaments are attached to polystyrene beads, and these myosin beads are deposited on the cytoplasmic face of a dissected Nitella axillaris cell. These cells contain rows of chloroplasts attached to the inner face of the cell membrane. Bundles of actin filaments of uniform polarity lie on these chloroplast rows (8). Myosin beads that come into contact with the actin bundles move unidirectionally over long distances in the same direction as cytoplasmic flow in the unopened cell. The rates of movement of the beads can be determined by noting position and time elapsed at several points along the bead path. The Nitella-based assay yields reproducible and quantitative rate data with relatively small samples of purified myosin. Nonetheless, this assay is flawed because it depends on the biochemically undefined actin cables of the Nitella cell, which are stabilized by unknown factors and may be contaminated by components of the Nitella cytoplasm. Thus, we have worked to replace the Nitella-based assay with a movement assay using only purified proteins.The Nitella-based assay demonstrated a principle that we have exploited in designing movement systems composed of purified proteins. Although the myosin filaments are bipolar and are bound to the beads in random orientations (9), the beads move in only a single direction-that defined by the polarity of the actin filaments. This suggests that only those myosin cross-bridges oriented properly can interact productively with an actin filament. Those not oriented correctly presumably cannot recognize their binding site on actin. Thus, the direction of movement is defined by the polarity of the actin...