Purpose: Neuromuscular Electrical Stimulation (NMES) is commonly used in neuromuscular rehabilitation protocols, and its parameters selection substantially affects the characteristics of muscle activation. Here, we investigated the effects of short pulse width (200 µs) and higher intensity (short-high) NMES or long pulse width (1000 µs) and lower intensity (long-low) NMES on muscle mechanical output and fractional oxygen extraction. Muscle contractions were elicited with 100 Hz stimulation frequency, and the initial torque output was matched by adjusting stimulation intensity. Methods: Fourteen able-bodied and six spinal cord-injured (SCI) individuals participated in the study. The NMES protocol (75 isometric contractions, 1-s on–3-s off) targeting the knee extensors was performed with long-low or short-high NMES applied over the midline between anterior superior iliac spine and patella protrusion in two different days. Muscle work was estimated by torque–time integral, contractile properties by rate of torque development and half-relaxation time, and vastus lateralis fractional oxygen extraction was assessed by Near-Infrared Spectroscopy (NIRS). Results: Torque–time integral elicited by the two NMES paradigms was similar throughout the stimulation protocol, with differences ranging between 1.4% (p = 0.877; able-bodied, mid-part of the protocol) and 9.9% (p = 0.147; SCI, mid-part of the protocol). Contractile properties were also comparable in the two NMES paradigms. However, long-low NMES resulted in higher fractional oxygen extraction in able-bodied (+ 36%; p = 0.006). Conclusion: Long-low and short-high NMES recruited quadriceps femoris motor units that demonstrated similar contractile and fatigability properties. However, long-low NMES conceivably resulted in the preferential recruitment of vastus lateralis muscle fibers as detected by NIRS.