1. Experiments were carried out on an in vitro preparation of the lamprey brainstem isolated together with intact labyrinths. Responses of reticulospinal neurons from different brainstem reticular nuclei (mesencephalic, MRN; anterior rhombencephalic, ARRN; middle rhombencephalic, MRRN; and posterior rhombencephalic, PRRN) to rotation of the preparation (0 degrees-360 degrees) either in the sagittal plane (pitch tilt, or nose up-down movement) or in the transverse plane (roll tilt, or left-right inclination) were recorded. 2. Responses to roll tilt were qualitatively similar in all nuclei: contralateral side down tilt (in relation to the location of the neuron in the brain) caused an activation of reticulospinal neurons. The angular thresholds for activation differed, however, between nuclei as well as the angle at which the maximal activity occurred. The maximal response for MRN was at 45 degrees, for MRRN and PRRN at 90 degrees, for ARRN at 180 degrees. Thus, the zones of spatial sensitivity differed in different nuclei, and they covered the whole range of possible inclinations in the transverse plane. 3. Responses to pitch tilt were not uniform in the different nuclei. MRN neurons responded preferentially in the range of 45 degrees-90 degrees nose-up inclinations, but a proportion of the cells responded in the range of 45 degrees-90 degrees nose-down inclinations. The ARRN neurons had their maximal response when the brain was turned to a dorsal side-down position (180 degrees). In the MRRN, three subgroups of neurons could be distinguished, the first responding at around 90 degrees nose-down, the second responding at around 90 degrees nose-up and the third responding in both zones. However, the activation in the nose-up zone was less robust: responses in this zone were present only in approximately one half of the experiments. Finally, the PRRN neurons were found to be very heterogeneous, with their zones of sensitivity being distributed throughout the whole space (0 degrees-360 degrees). Thus, also in the sagittal plane, the zones of spatial sensitivity in the different nuclei covered the whole range of possible inclinations. 4. Long-term recording of MRRN neurons having the zone of sensitivity around 90 degrees nose-up showed that this response was rather unstable. Its amplitude varied considerably and could disappear with time to reappear later. These results, together with the fact that in a part of the experiments the MRRN neurons responded only in the 90 degrees nose-down zone (see above), leads us to suggest that the system of spatial orientation can dynamically re-organize.(ABSTRACT TRUNCATED AT 400 WORDS)