It is unknown how knee osteoarthritis pain affects joint power distribution while cycling. The study purposes were to (1) investigate if seat height, workload and any difference in hip or knee extensor strength affected asymmetry of hip, knee and ankle joint power during cycling;​ and (2) determine the relationship between knee osteoarthritis pain asymmetry and joint power asymmetry at the hips, knees, ankles and total leg. Asymmetry was the difference between dominant and non-dominant legs. Twenty-six participants (13 female, 13 male) with (n = 21) and without (n = 5) symptomatic knee osteoarthritis participated [age 64.3 (7.3) y, body mass index 27.0 (4.1) kg/m²]. Participants completed six cycling bouts at three seat heights (20°, 30°, 40° minimum knee flexion angle) and two workloads (40 W, 75 W) on a stationary bike. Self-reported pain was recorded for each knee before the first bout and after each bout. Three-dimensional kinematics and kinetics were collected synchronously with motion capture and instrumented pedals. A greater workload was associated with greater hip power asymmetry (p < 0.01);​ otherwise, seat height and workload did not affect power asymmetry (p > 0.05). Relationships were found between knee pain asymmetry and hip, ankle and total leg power asymmetry (p < 0.01), but not knee (p > 0.05). The hip, ankle and total leg with the more painful knee produced less power than the opposite side. The more painful knee cannot be assumed to produce less power than the contralateral side. These findings show that, at low workloads, clinicians can adjust seat height to patient preference without affecting joint power production during cycling.