This study investigated lower-limb kinematics to explain the techniques used to achieve high levels of sprint start performance. A cross-sectional design was used to examine relationships between specific technique variables and horizontal external power production during the block phase. Video data were collected (200 Hz) at the training sessions of 16 sprinters who ranged in 100 m personal best times from 9.98 to 11.6 s. Each sprinter performed three 30 m sprints and reliable (all intraclass correlation coefficients, ICC(2,3) ≥ 0.89) lower-limb kinematic data were obtained through manual digitising. The front leg joints extended in a proximal-to-distal pattern for 15 sprinters, and a moderate positive relationship existed between peak front hip angular velocity and block power (r = 0.49, 90% confidence limits = 0.08-0.76). In the rear leg, there was a high positive relationship between relative push duration and block power (r = 0.53, 90% confidence limits = 0.13-0.78). The rear hip appeared to be important; rear hip angle at block exit was highly related to block power (r = 0.60, 90% confidence limits = 0.23-0.82), and there were moderate positive relationships with block power for its range of motion and peak angular velocity (bothr = 0.49, 90% confidence limits = 0.08-0.76). As increased block power production was not associated with any negative aspects of technique in the subsequent stance phase, sprinters should be encouraged to maximise extension at both hips during the block phase.