Background: We present computational analysis and inference tools to assess actomyosin interaction in the motility assay. Methods/Results: Analysis: Individual actin filaments are automatically detected and tracked by area and position similarity in consecutive video frames. Filament velocities (V) are calculated as distance traveled over elapsed time (T). For individual filaments, the software measures filament length (L), velocity (V), number of immotile frames. For the overall video, the breakage of filaments is monitored. Velocity averages were found to be linearly related to values obtained manually using Scion Image v.4.02 (NIH Image) (proportionality factor:0.96±0.10, from 8 videos). Model: A model was developed that incorporates actomyosin binding sites with independent kinetics (2 states:attached/detached, Poisson escape statistics: mean attached time τon and mean detached time τoff, binding site distance (B)). The mean filament velocity is V(L) = Vmax[1-Poff⊥(L/B)], with Poff = τoff /(τon+τoff), the probability for a single binding site to be idle. Numerical analysis of the variance suggests that the standard deviation (σ) over mean V is σ/V = C/sqrt(T∗L), C=constant. Fitting: Fitting mean velocity V(L) and C to data allows the inference of Vmax, Poff, τon. We simulate filament data for known parameters, reverse fit these parameters, and for 6 simulated data sets (each holding 40 filaments) we find the estimated parameter values to deviate from the input by 6%, 20%, and 20%, respectively. Benefits/Implications: Our automated analysis software provides: 1.High filament counts 2.Detailed and reliable tracking of single filament motion 3.Flexibility to meet the specific needs of the experimental operator. Small amounts of motility data suffice to fit three important kinetic parameters. The model predicts increased variance for slow myosin types, as is observed for smooth muscle myosin. NSERC RGPIN217457-2010.