Objectives. Determine whether a rehabilitation program targeting functional motor recovery of persons with poststroke hemiparesis improved motor coordination. Methods. A subgroup of 20 persons with poststroke hemiparesis (n = 11 in intervention and n = 9 in control group) was investigated from within a larger randomized controlled single-blind clinical trial of 100 patients. Motor coordination was measured using a pedaling task, and subjects in the intervention group pedaled during an intensive broad-based home exercise program that targeted flexibility, strength, balance, and endurance. Coordination variables based on paretic leg pedal forces and EMG of 4 thigh muscles were measured while pedaling pre- and postintervention. Results. Despite extensive pedaling practice, up to 30 half-hour sessions that were progressively more intense, there was no effect (P > 0.05) of the intervention on percent of total work done by the paretic leg, quantitative measures of EMG, or pedaling speed. However, walking speed was improved and pedaling and walking faster were associated after the intervention. Conclusions. There is no evidence of improved locomotor coordination postintervention. The increased walking and pedaling speed was likely achieved by a more proficient use of the same impaired pattern without EMG timing changes, likely because of increased strength and endurance postintervention. A more task-specific intervention may be required to improve coordination, consistent with principles of use-dependent plasticity.