The aim of this study was to investigate human adaptations to fatigue induced by track sprint repetitions. Eight male sprinters were asked to run 4 × 100 m as quickly as possible with 3 min of recovery. Subjects were filmed (50 Hz) in order to measure stride length and frequency. Velocity was measured by means of radar (250 Hz) while contact and flight times were registered wirelessly by two pressure sensors (400 Hz) embedded under the insole of the subjects’ shoes. Contact and flight times were used to calculate stiffness. In addition, blood samples were taken prior to warm-up, 1 min after each 100-m sprint and every 2 min after the last repetition until a lactate peak ([BLa]) was reached. [BLa] did not affect mechanical and stride parameters. Inter-series ANOVA showed that velocity decreased significantly (−3.55%) between Repetition 1 (8.18 ± 0.29 m s⁻¹) and Repetition 4 (7.89 ± 0.42 m s⁻¹), while [BLa] increased from 6.74 ± 1.15 to 13.58 ± 1.48 mmol l⁻¹ (p < 0.05). The first main result was that leg stiffness remained constant until Repetition 3 and then dramatically increased at Repetition 4, whereas vertical stiffness remained constant throughout all four repetitions. This behavior could be considered to be a neuromuscular adaptation to fatigue used by skilled athletes. The second main result was that velocity decreased during the second phase (30–80 m) of the entire 100 m. In addition, a PCA revealed three different sprint profiles explaining 88.2% of the total variance: the contact-time-pattern (39.46%), force-pattern (27.96%) and stride-pattern (20.77%). Two different motor signatures were identified with fatigue. In the first, athletes switch from the key variable to another when exhausted without changing their motor behavior (during Repetition 3 and/or Repetition 4). In the second, athletes do not change their motor behavior with fatigue.