Sensitivity of lateral motions relative to vertical motions were determined and compared to predictions provided by ISO 2631. Two experiments were executed where lateral and vertical motions were applied consecutively or simultaneously and where the magnitude of a single- or dual-axis test signal was adjusted until it was judged as equivalent to a preceding single-axis reference motion of the same frequency. Experiment 1: References consisted of vertical sinusoidal motions presented with 1.6-12.5 Hz and weighted accelerations of a(zw) = 0.3, 0.6 and 1.2 m s(-2) r.m.s., single-axis test signals were lateral motions of the same frequency. 26 subjects (15 men, 11 women, 20-56 yr) participated in the experiments. Accelerations adjusted for lateral vibrations above 1.6 Hz were considerably lower than predicted suggesting that the weighting factors provided in ISO 2631 are incorrect. Experiment 2: References consisted of single-axis vertical or lateral sinusoidal motions presented with 1.6-12.5 Hz and a weighted acceleration of a(zw) = 1.25 m s(-2) r.m.s. The dual-axis test signals consisted of a constant fraction of the reference acceleration (10, 25, 50, 75, 90%) and a perpendicularly oriented adjustable component. 31 subjects (15 men, 16 women, 19-51 yr) participated in the experiments. Both experiments revealed that ISO 2631 is qualitatively valid, the weighting of lateral motions above 1.6 Hz, however, should be increased in order to meet the actual sensitivity particularly in case of multi-axis vibrations.