Aim: Intramuscular botulinum toxin-A (BoNT-A) injections reduce spasticity by blocking neurotransmission at the motor endplate (MEP). The goal of this study was to assess the reduction in spasticity achieved by injecting BoNT-A at different sites of the gracilis muscle. Method: Thirty-four gracilis muscles, in 27 children (10 females and 17 males, mean age of 8.6y [SD 2.5y]) with spastic cerebral palsy (unilateral and bilateral, Gross Motor Function Classification System [GMFCS] levels I-IV), were randomly assigned to one of two groups. In one group BoNT-A was injected proximally (at a site 25% of the distance from the pubic tubercle and the medial epicondyle) and in the other it was injected at the MEP zones (half of the dose was administered at 30% of this distance and half at 60%). Spasticity was assessed before and after BoNT-A injection using simultaneous measurements of surface electromyography (sEMG) and angular velocity during passive muscle stretch applied at different velocities. The primary outcome measure included the velocity-dependent change in average root mean square electromyography (RMS-EMG). Secondary outcome was assessed with the Modified Ashworth Scale (MAS) and Modified Tardieu Scale (MTS). Results: Spasticity decreased more in MEP-targeted muscles than in proximally injected muscles, as demonstrated by a larger reduction in average RMS-EMG values (p=0.04), though this difference was not found with the MAS or MTS. Interpretation: The results suggest that BoNT-A injection of the gracilis at sites with a high concentration of MEPs is effective at reducing spasticity. These preliminary findings should be confirmed by larger studies. In the case of long muscles, such as the gracilis, the injection site is important.