Interacting with a continuously changing environment may necessitate cancelling a programmed movement, such as a step, after an environmental change to avoid errors and possible falls. This cancellation is successful if it occurs during the few milliseconds from the Anticipatory Postural Adjustments (APA) that precede the step onset, a period in which the center of mass (CoM) and the center of pressure (CoP) can still be aligned. Here we studied the inhibition of steps by the means of Stop Signal Task (SST), a behavioral task matched with a theoretical model developed to study the behavioral and physiological correlates of simple response inhibition [1], e.g., pressing a button. Aim of the present work was investigating if and how the inhibition of a complex multi-effectors movement as in step, accomplishes the assumptions of the model developed for simpler movements and if it is they are reflected in the time evolution of CoM and CoP. 12 healthy subjects (age: 33.3±7.1 y; weight: 69.3±17 kg; height: 169.7±8.2 cm) were required to initiate 300 walks from a stationary position (Go trials; GT) in response to the onset of an ahead pointing arrow (Go signal) on a pc monitor, and to interrupt the movement when in 30% of the trials (Stop trials; ST), a road stop signal (Stop signal) replaced the arrow after a variable time, determined using a behavioral tracking staircase algorithm [2]. A stereo-photogrammetric motion analysis system (SMART-DX 6000: BTS) and eight dynamometer platforms (Kistler 9286B; Kistler) were used to detect the kinematic and dynamic parameters of gait initiation. Ad-hoc analysis algorithms were developed by the MATLAB software (R2019b 9.7; MathWorks) to track the time evolution of CoM and CoP trajectories and capture the related reaction times (RT) for GT, error stop trials (eST), and correct stop trials (cST), and to conduct the statistical analysis. Fig. 1 displays the CoM (left upper panels) and CoP (left lower panels) trajectories in GT, eST, and cST, and the computed corresponding RT (rightmost panels). A repeated measures ANOVA and post-hoc tests detected (p<0.001), for both CoM and CoP, a significantly slowest RTs in cST (CoM=375±53ms; CoP=277±64ms), faster RT in GT (CoM=335±62ms; CoP=253±70ms), and the even faster in RT during the eST (CoM=296±61ms; CoP=213±64ms). By tracking the time evolution of CoM and CoP in a step version of the SST we obtained, during APA, a measure of RT, not only in GT and eST trials, but even in cST, a parameter hidden in the typical experimental setup. With these variables, we detected that the outcome of our version of SST complied with the main assumption of the race model: a movement is successfully cancelled when a stop process wins the race being faster than a go process. Accordingly, here we observed the eST were the trials approaching the race with significantly faster RT than cST. These results provide evidence that our version of SST is a valid tool for studying inhibition of complex movements and the associated kinematic variables.