Stroke is a major cause of disability in the US, leaving most survivors with abnormal motor function, often resulting in hemiparetic gait. Quality of gait in this population has been well characterized by measures of velocity, though velocity alone provides only a general index of functional mobility. Researchers and clinicians should continue to develop a complete characterization of behavioral alterations, but should also work to identify the neuromechanical processes underlying the loss and recovery of locomotor function. We hypothesized that selected foot center of pressure (CoP) measures would show a predictive relationship with hemiparetic gait velocity, a standard clinical metric of locomotor capability, thus providing insight into the relationship between neuromotor control and function. Thirty-three chronic stroke survivors (67+/-10 years) were evaluated during walking at a self-selected speed. Patients wore pressure sensitive shoe insoles and completed 15 steady-state gait cycles over an instrumented gait mat. CoP parameters and walking velocity were measured during each cycle. Multiple regression analyses were used to model all combinations of CoP and interlimb symmetry parameters as predictor variables of gait velocity. Eleven CoP and symmetry parameters were selected in a final regression model, and provided a robust prediction of hemiparetic gait velocity (R(adj)(2)=0.90) in this group of chronic stroke patients. These results indicate that bilateral foot CoP measures, especially those representing variability of foot CoP control, not only index locomotor function, but may also have the potential to provide information about the underlying control properties of the stroke-injured neuromuscular system.