Velocity variation in swimming is commonly assessed using quantification methods that fail to consider the movement mechanical nature in the aquatic environment. The current study proposes the relative power to velocity variation method to evaluate the velocity variation effect in efficiency, accounting the swimmers hydrodynamic drag into the calculation. Twenty regional level swimmers (12 males) performed three 25 m front crawl trials (one at 88 and other at 100 % of their maximum velocity, and another at maximum pace using the velocity perturbation method). Mean, maximum and minimum velocities, stroke rate, length and index, and indexes of coordination and synchronization were obtained for each cycle. The power to overcome drag, power to velocity variation, total mechanical power, intracycle velocity variation (assessed by the coefficient of variation) and difference between maximum and minimum absolute velocities were also computed. Power regressions were performed between mean velocity, standard deviation and the absolute and relative power to velocity variation. Results showed that the absolute and relative power to velocity variation values increased (a = 2.82, b = 1) and decreased (a = 0.07, b = −2.00, respectively) with the mean velocity increment, while both rose with mean velocity standard deviation (a = 142.31, b = 2.00 and a = 1.14, b = 2.00, respectively). The relative power to velocity variation method offers an advanced understanding of front crawl efficiency and enables predicting its effect on overall swimming performance.