The purpose of the present study was to test the traditional assumption that sea sickness is uniquely provoked by heave motion characteristics, pitch and roll movements being ineffective. In an experi-ment with a ship motion simulator subjects were exposed to pitch, and roll motions in combination with rather weak heave motions that have no motion sickness inducing potential. Very high levels of motion sickness were observed (with a motion sickness rating scale) in almost 50% of our subjects. In three control experiments it was shown that these heave motions, when presented separately, have no motion sickness inducing potential indeed, and that pitch and roll motions presented alone or in combination with each other, have only a very small motion sickness inducing potential. These results indicate that pitch and roll when combined with small heave motions which in themselves are not sickness provoking produce more motion sickness than claimed by the classical models. This suggests that, in models on motion sickness, pitch and roll should be combined in a non-linear fashion with heave, and that such models will remain rather crude if they do not include a description of the vestibular contribution to motion sickness.