Limited evidence showed that higher workload increases knee forces without effects from changes in pedalling cadence. This study assessed the effects of workload and cadence on patellofemoral and tibiofemoral joint forces using a new model. Right pedal force and lower limb joint kinematics were acquired for 12 competitive cyclists at two levels of workload (maximal and second ventilatory threshold) at 90 and 70 rpm of pedalling cadence. The maximal workload showed 18% larger peak patellofemoral compressive force PFC (large effect size, ES) than the second ventilatory threshold workload (90 rpm). In the meantime, the 90-rpm second ventilatory threshold was followed by a 29% smaller PFC force (large ES) than the 70-rpm condition. Normal and anterior tibiofemoral compressive forces were not largely affected by changes in workload or pedalling cadence. Compared to those of previous studies, knee forces normalized by workload were larger for patellofemoral (mean = 19N/J; difference to other studies = 20-45%), tibiofemoral compressive (7.4N/J; 20-572%), and tibiofemoral anterior (0.5N/J; 60-200%) forces. Differences in model design and testing conditions (such as workload and pedalling cadence) may affect prediction of knee joint forces.