Fifteen healthy men performed a 5-week training program comprising four sets of seven unilateral, coupled concentric-eccentric knee extensions 2-3 times weekly. While eight men were assigned to training using a weight stack (WS) machine, seven men trained using a flywheel (FW) device, which inherently provides variable resistance and allows for eccentric overload. The design of these apparatuses ensured similar knee extensor muscle use and range of motion. Before and after training, maximal isometric force (MVC) was measured in tasks non-specific to the training modes. Volume of all individual quadriceps muscles was determined by magnetic resonance imaging. Performance across the 12 exercise sessions was measured using the inherent features of the devices. Whereas MVC increased (P < 0.05) at all angles measured in FW, such a change was less consistent in WS. There was a marked increase (P < 0.05) in task-specific performance (i.e., load lifted) in WS. Average work showed a non-significant 8.7% increase in FW. Quadriceps muscle volume increased (P < 0.025) in both groups after training. Although the more than twofold greater hypertrophy evident in FW (6.2%) was not statistically greater than that shown in WS (3.0%), all four individual quadriceps muscles of FW showed increased (P < 0.025) volume whereas in WS only m. rectus femoris was increased (P < 0.025). Collectively the results of this study suggest more robust muscular adaptations following flywheel than weight stack resistance exercise supporting the idea that eccentric overload offers a potent stimuli essential to optimize the benefits of resistance exercise.