Rhythmic spinal circuits and reflexes can function to control and reinforce fundamental locomotor muscle activity patterns. An understanding of functional locomotor synergies provides insight about the functional organization of neuronal elements. Although these synergies are assumed to be bilaterally symmetrical, unequal motor output and functional locomotor asymmetries commonly occur, as in gait over irregular surfaces or in persons with musculoskeletal injury or neural pathology. In these studies, asymmetrical pedaling, where one limb produces significantly less crank work than the contralateral limb, can theoretically be controlled most simply by modulating only the amplitudes of muscle activity. However, because they span two joints, a phasing response in biarticular muscles may be a preferred coordination strategy to produce asymmetrical pedaling. It was hypothesized the asymmetrical task would elicit modulation of biarticular muscle activity phasing and uniarticular muscle activity amplitude, and that biarticular muscle activity would phase-shift differently in forward versus backward pedaling. Healthy, young subjects were instructed to pedal at a moderate cadence (40 rpm) and bilateral workload (130 Joules/cycle) while emulating torque profiles of bilaterally unequal limb crank work with real-time feedback during both forward and backward pedaling. Results showed contralateral limb increased work was achieved with increased amplitudes in all muscles while target limb reduced work was associated with reduced amplitude in the uniarticular vastus medialis. Target limb biarticular muscles phase advanced proportional to the work reduction in forward pedaling, but exhibited a “blended” strategy of both phasing (a discrete switch to prolonged excitation) and amplitude modulation in backward pedaling. A proposed model suggests the biarticular muscle coordination depends on both uniarticular muscle amplitude and the initial relative phasing of biarticular and uniarticular muscle activity in the task. Accordingly, conceptual models of neural control may need to incorporate bilaterally independent biarticular muscle phasing responses to changes in the locomotor task.