This study aimed to establish the effect of cycling mode and cadence on torque, external power output, and lower limb muscle activation during maximal, recumbent, isokinetic cycling. After familiarisation, twelve healthy males completed 6 × 10 s of maximal eccentric (ECC) and concentric (CON) cycling at 20, 40, 60, 80, 100, and 120 rpm with five minutes recovery. Vastus lateralis, medial gastrocnemius, rectus femoris, and biceps femoris surface electromyography was recorded throughout. As cadence increased, peak torque linearly decreased during ECC (350-248 N·m) and CON (239-117 N·m) and peak power increased in a parabolic manner. Crank angle at peak torque increased with cadence in CON (+13°) and decreased in ECC (-9.0°). At all cadences, peak torque (mean +129 N·m, range 111-143 N·m), and power (mean +871 W, range 181-1406 W), were greater during ECC compared to CON. For all recorded muscles the crank angle at peak muscle activation was greater during ECC compared to CON. This difference increased with cadence in all muscles except the vastus lateralis. Additionally, peak vastus laterallis and biceps femoris activation was greater during CON compared to ECC. Eccentric cycling offers a greater mechanical stimulus compared to concentric cycling but the effect of cadence is similar between modalities. Markers of technique (muscle activation, crank angle at peak activation and torque) were different between eccentric and concentric cycling and respond differently to changes in cadence. Such data should be considered when comparing between, and selecting cadences for, recumbent, isokinetic, eccentric and concentric cycling.