Background Some skilfulfast activities require control of trajectory and force to fulfill the task. It is known that voluntary reactions can be speeded up by a startling auditory stimulus (SAS) delivered at the same time as the imperative signal (IS), a phenomenon termed Start-React (Valls-Sole et al., 1999). The aim of the present study was to examine how the performance of a task that demands accuracy, mainly in terms of velocity during the trajectory or force control when holding an object, is changed by a SAS timed to occur either at the time of IS or at a pre-determined time interval when movement has already begun and motor control of accurate reaching is expected. Methods Nine subjects performed an accurate elbow extension between two spots placed at an angular distance of 30° when holding with their right hand a pen that monitored, through two strain gauge systems, the force of the subject’s fingers during the hold, and the force at which pen-tip was exposed with table contact. According to block of trials, this end spot had three possible diameters: 5, 10, and 20 mm. Kinematic and force parameters were measured for three conditions, control, with SAS at IS or SAS delivered when movement has already started. Results In SAS trials, there were no evident differences with respect to control trials in the general outline but there were differences in the timing of the events. In SAS-IS trials time to peak velocity was 166 ms (SD = 41 ms), time to peak force was 95 ms (SD = 43 ms) both shorter than in control or SAS delivered when movement began. Peak velocity (315 ± 56 o/s) and peak force (8 ± 2 arbitr. units) were larger for SAS-IS trials than for control or SAS delivered when movement started. Conclusions Startle can advance a task that requires accuracy with velocity and force modulations, but when task onset is already launched there are not evident changes in velocity or force parameters, that seem already prepared.