Introduction Surface electrical stimulation can be performed by using electrodes placed over the muscle (muscle stimulation) or over a major motor nerve (nerve stimulation). The purpose of the present study was to compare electrical (M-wave) and mechanical (twitch) properties of the quadriceps muscle evoked by single pulse delivered via muscle vs. nerve stimulation of increasing intensity. We hypothesized that motor unit (MU) recruitment will differ between the two stimulation modes. Methods Twenty two healthy subjects (16 men and 6 women, 29±2 yrs) were recruited for this study. They were seated with a knee angle of 90°. Electrodes were placed over the femoral nerve (diameter: 5 cm) and on the gluteal fold (10 x 5 cm) for nerve timulation. For muscle stimulation, large electrodes (15 x 9 cm) were located 5-10 cm below the inguinal crease and 5-10 cm above the superior border of the patella. Peak twitch amplitude (Pt) and time to peak twitch (TPT) as well as M-wave amplitude from the vastus lateralis (VL) and vastus medialis (VM) muscles were obtained for stimulus of increasing intensity using muscle and nerve stimulation. The resulting recruitment curves and discomfort (visual analogue scale from 0 to 10) associated to maximal stimulation intensity (defined as the intensity where a plateau in M-wave and twitch responses was obtained) were analysed. Results Muscle and nerve stimulations resulted in similar Pt (57.3 ± 16.8 N vs. 60.6 ± 15.8 N, P>0.05) and similar M-wave amplitude for VL and VM muscles at maximal stimulation intensity. However, some differences between the two stimulation modes were found at submaximal stimulation intensities for both mechanical and electrical responses. TPT de- creased (P<0.05) with increased stimulation intensity for both stimulation modes with no difference between stimulation modes. At max- imal stimulation intensity, muscle stimulation resulted in a lower discomfort than nerve stimulation (3.2 ± 1.4 vs. 5.6 ± 1.6; P<0.001). Con- clusion The shape of the recruitment curves showed some discrepancies between the two stimulation modes, suggesting that MU acti- vation pattern differed. However, the reduction in TPT with increasing intensity suggest that Mus were orderly recruited with both stimula- tion modes, which is in contrast with the hypothesis of MU random activation with muscle stimulation (Bickel et al. 2011). Maximal M-wave amplitude and Pt were similar with muscle vs. nerve stimulation, suggesting that the use of muscle stimulation with large electrodes might be an alternative to nerve stimulation as it results in less discomfort.