At a constant power output, cyclists prefer to use a higher cadence than those that minimise metabolic cost. The neuromuscular mechanism underpinning the preferred higher cadence remains unclear. Purpose. The aim of this study was to investigate the effect of cadence on joint level work and vastus lateralis (VL) fascicle mechanics while cycling at a constant, submaximal, power output. We hypothesised that preferred cycling cadence would enhance the power capacity of the VL muscle when compared to a more economical cadence. Furthermore, we predicted that the most economical cadence would coincide with minimal total electromyographic activity from the leg muscles. Methods. Metabolic cost, lower limb kinematics, joint level work, VL fascicle mechanics, and muscle activation of the VL, rectus femoris, biceps femoris, gastrocnemius medialis and soleus muscles were measured during cycling at a constant power output of 2.5 W/kg and cadences of 40, 60, 80 and 100 revolutions per minute (RPM). A preferred condition was also performed where cadence feedback was hidden from the participant. Results. Metabolic cost was lowest at 60 RPM, but the mean preferred cadence was 81 RPM. The distribution of joint work remained constant across cadences, with the majority of positive work being performed at the knee. The preferred cadence coincided with the highest VL power capacity, without a significant penalty to efficiency, based on fascicle shortening velocity. Conclusions. Cycling at a higher cadence is preferred to ensure that the muscle's ability to produce positive power remains high. Further investigations are required to examine what feedback mechanism could be responsible for the optimisation of this motor pattern.