This chapter aims to improve flocking control for a group of nonholonomic robots. It introduces a new flocking control algorithm with potential-based flocking as its foundation. By incorporating Leader's Backstepping algorithm into the flocking strategy, an improved flocking performance is obtained, which leads the flock to the target point swiftly in a smoothed trajectory. Simulations in this chapter test and verify the effectiveness of the algorithm, in which key parameters' influences on system performance are discussed.