Gaze stabilizing eye and head responses compensate more effectively for low-frequency rotational motion when such motion stimulates the otolith organs, as during earth-horizontal axis rotations. However, the nature of the otolith signal responsible for this improvement in performance has not been previously determined. In the current study, we used combinations of earth-horizontal axis rotational and translational motion to manipulate the magnitude of net linear acceleration experienced by pigeons, under both head-fixed and head-free conditions. We demonstrate that phase enhancement of eye and head responses to low frequency rotational motion was causally related to the magnitude of dynamic net linear acceleration, not the gravitational acceleration component. We also show that canal-driven and otolith-driven eye responses were both spatially and temporally appropriate to combine linearly, and that a simple linear model combining canal- and otolith-driven components predicted eye responses to complex motion that were consistent with our experimental observations. However, the same model did not predict the observed head responses, which were spatially but not temporally appropriate to combine according to the same linear scheme. These results suggest that distinct vestibular processing substrates exist for eye and head responses in pigeons, and that these are likely different from the vestibular processing substrates observed in primates.