The contact time (CT) and the ratio between the duration of braking (BP) and propulsion phase (PP) during the support are two mechanical parameters that are relevant for the performance in sprinting. Several different kinematic methods have been used to estimate the BP-PP transition point, with some disagreements among studies. The purpose of this study was to compare three criteria to individuate that point: the maximum knee flexion (KNEEFLEX), the maximum fall of the COM (COMVERT) and the horizontal COM acceleration (COMHORIZ). It was hypothesized that these three events would take place in different instants, representing different moments of the step cycle. The kinematic analysis was performed through stereophotogrammetry on seven male middle-level sprinters (mean + or - SD height: 180 + or - 5 cm; body mass: 75 + or - 11 kg; personal best on 100 m: 10.96 + or - 0.45 s). The COMVERT expressed as percentage of CT (mean + or - SD 31.73 + or - 8.29%) was on average strongly anticipated compared to both KNEEFLEX (45.07 + or -3.60%, p < 0.01) and COMHORIZ (56.86 + or - 8.56%, p < 0.01). The approximately 12% difference between KNEEFLEX and COMHORIZ was also statistically significant (p < 0.01). The 40-60% relative proportion of BP and PP considered to be optimal in the literature, was reflected in this study only by the KNEEFLEX criterion. Given the importance of BP time as a reference parameter to assess the training status and efficacy of training interventions, it looks necessary both from a scientific and applied perspective to consider a standardized criterion to identify the BP-PP transition point.