This study evaluated the retention of a cycling kinetic pattern using two different feedback schedules and evaluated the potential for feedback dependency in a continuous-task learning environment. Eighteen inexperienced cyclists rode a racing bicycle mounted to a fixed-fork Velodyne Trainer, with pedal forces monitored by dual piezoelectric transducers. Subjects received right-pedal shear force feedback and a criterion pattern emphasizing "effective" shear. Concurrent feedback (CFB) subjects received concurrent feedback 140 ms after the completion of every other revolution, while summary feedback (SFB) subjects received averaged feedback between trials. All subjects performed 10 retention trials without feedback 1 week later. Both groups improved significantly during practice, and performance decay in retention was negligible. Group differences during all phases were not significant. High CEB group proficiency in retention indicated that the detrimental aspects of frequent feedback were not significant. High SFB proficiency in retention suggests that large changes in kinetic patterning are achievable with relatively few feedback presentations.