The statistics of heavy meromyosin (HMM) driven actin filament paths in vitro, and thermal fluctuations of actin filaments suspended in a pseudo 2D-space in solution, can be described by the cosine correlation equation (CCE): <cos(qs)-q0))> = exp(-s/[2*Lp]). Here, q0) and qs) represent tangent angles at distance 0 and s, respectively from one filament end (in solution) or from the starting point of the path. The quantity Lp is the persistence length (proportional to flexural rigidity) of the filament/path. In vitro motility assay (IVMA) studies (27-29oC) were performed along with studies of actin filaments suspended between two cover-slips in solution. Fits to the CCE gave LP = 16.5 5 1.7 mm (mean 5 95 % confidence interval) and 11.1 5 0.6 mm for phalloidin stabilized filaments in solution and propelled by HMM, respectively. In contrast, phalloidin free actin filaments (NHS-rhodamine labeled) exhibited similar LP in solution 10.1 52.1 mm and during HMM propulsion (9.8 5 0.9 mm). The filament paths were modeled using a Monte-Carlo approach updating angular changes in sliding direction at short time intervals (dt) assuming 1. lateral displacements due to cross-bridge forces and 2. thermal fluctuations of the leading filament end. The results suggest that > 3nm average lateral displacement during each actomyosin interaction would reduce LP by > 30 % compared to that of filaments without HMM. The findings are consistent with the following ideas: 1. Actin filaments exist in two different flexural rigidity states, one favored by myosin binding and the other by phalloidin stabilization, 2. Changes in actin filament flexural rigidity is not required for motion generation. 3. The myosin cross-bridges produce minimal lateral movements (< 3 nm) during the power-stroke.