It has been suggested that, after a passive linear acceleration of a seated subject which resembles a small, rear-end car impact, sensory information from proprioceptive, vestibular, and visual systems elicit stabilizing neck muscular responses. These neck muscular responses are presumably reflex based and are modified with the magnitude of the perturbation. A key issue that remains is to determine whether the neck and head postural responses can be modulated by a previous experience of the acceleration and not only by the magnitude of the acceleration. This question is of interest because, contrary to cadaver studies, one could expect that humans apprehending a rapid trunk acceleration would adopt a bracing behavior to minimize head movements. The aim of the present experiment was to verify whether neck-muscle activities can be modulated when prior knowledge about whole-body acceleration onset, direction, and magnitude are unknown compared with when only acceleration onset is unknown. Nine seated subjects were submitted to 11 imposed, forward linear accelerations (1.1g). For the first trial, subjects were completely unaware of the platform acceleration characteristics (onset, direction, amplitude, and acceleration magnitude). For the subsequent ten trials, subjects knew they would be submitted to a forward linear acceleration, but the onset of the acceleration was unknown. Head kinematics and EMG responses of the neck muscles to the first perturbation were similar for all subjects (6.2° head extension, EMG activity starting from 55 to 72 ms after platform onset). Following the first trial, however, all subjects showed a decreased neck EMG activity. Moreover, subjects responded in one of two ways across trials:​ one group of subjects (n=5) maintained a constant head angular position and velocity, whereas the other group (n=4) showed an increased head angular position (up to 12.6°) and velocity. This suggests that the first perturbation trial revealed a completely reactive response. After this initial trial, the responses observed may present a mixture of feedforward and feedback control. It is likely that whiplash injuries occur under conditions resembling those observed for the first trial only. If this is the case, the behavior for the following trials cannot be representative of injury mechanisms occurring in whiplash-like motion. Altogether, our results strongly suggest that, following repeated trunk linear accelerations of a constant magnitude, the nervous system prefers to minimize muscle stress instead of adopting a bracing strategy.