While local vibration (LV) has been recently proposed as a potential modality for neuromuscular conditioning, no practical recommendations to optimize its effects have been published. Because changes in corticospinal excitability may reflect at which degree the neuromuscular function is modulated during LV exposure, this study investigated the effects of muscle length and vibration site on LV-induced on motor evoked potentials (MEPs) changes. Twenty-one subjects participated in a single session in which MEPs were evoked on the relaxed knee extensors (KE) during three conditions, i.e., no vibration (CON), muscle (VIBMU), and tendon vibration (VIBTD). Three muscle lengths were tested for each condition, i.e., short/intermediate/long KE muscle length. Both VIBMU and VIBTD significantly increase MEPs compared to CON. Higher increases (P < 0.001) were found for VIBTD compared to VIBMU for vastus lateralis (mean increases of the three angles: +241% vs.+ 148%), vastus medialis (+273% vs. + 180%) and rectus femoris muscles (+191% vs. +141%). The increase in MEPs amplitude was higher (p < 0.001) at an intermediate (mean pooled increase for VIBTD and VIBMU: +265%, +290%, and +212% for VL, VM, and RF, respectively) compared to short (+136%, + 144%, and + 127%) or long (+ 184%, + 246% and + 160%) muscle lengths. These results suggest that LV should be applied to the tendon at an intermediate muscle length to optimize the acute effects of LV on the KE neuromuscular function.