To the best of our knowledge, no study has yet analyzed the most effective backstroke start variant based on the most recent rules modification endorsed by the Federation Internationale de Natation particularly regarding the neuro-muscular activation involved in the force production and stabilization of the body. PURPOSE: To analyze the electromyographic (EMG) activity of the propulsive and stabilizing muscles during backstroke starts performed with the feet submerged (BSFI) and with the feet above the water surface (BSFE). METHODS: Four high-level male swimmers (22.8±1.7 years, 75.9±8.9 kg, 1.78±0.06 m and 15.5±5.0 years of training experience) performed two sets of six maximal repetitions using BSFI and BSFE over a distance of 15m. Surface EMG signals of Deltoideus Anterior, Biceps Brachii, Triceps Brachii, Erector Spinae Longissimus, Gluteus Maximus, Rectus Femoris and Gastrocnemius Medialis were obtained. A qualitative video analysis was used to breakdown the backstroke start into five phases: hands-off, take-off, flight, entry and glide. The raw EMG signals were rectified, smoothed (10Hz), then integrated (IEMG) and normalized to maximum voluntary contractions and time. Descriptive statistics and paired Student's t test were used (p<0.05). RESULTS: During the hands-off phase it was registered higher IEMG for the Biceps Brachii and Erector Spinae Longissimus at BSFE comparing with BSFI (8.26±2.63% vs 3.85±1.37%, p<0.001; 3.85±1.27% vs 1.28±0.81%, p<0.001, respectively). No differences between start variants were observed in the take-off phase. For the 3rd phase (flight), BSFE presented higher IEMG on Deltoideus Anterior (1.63±0.78% vs 0.99±0.37%, p=0.01), Biceps Brachii (1.28±0.78% vs 0.93±0.78%, p<0.01), and Erector Spinae Longissimus (1.47±0.54 vs 0.63±0.56%, p<0.001). During the entry phase, the value of Erector Spinae Longissimus was lower for BSFI than for BSFE (2.11±1.85% vs 4.20±2.21%, p<0.001). The IEMG value for Biceps Brachii and Erector Spinae Longissimus at last phase (the glide) was higher for BSFE (1.28±0.31% vs 0.98±0.18%, p<0.01; 3.27±1.20 vs 1.97±1.43, p=0.01, respectively). CONCLUSIONS: The BSFE seems to require higher myoelectric activation than the BSFI from the stabilizing muscles of the back and the arms during the initial and middle part of the start movement.