This study uses frequency-domain techniques and stabilogram diffusion analysis (SDA) to investigate the effect of vibrotactile feedback during continuous multidirectional perturbations of a support platform. Eight subjects with vestibular deficits were subjected to two-axis pseudorandom surface platform motion while donning a multiaxis vibrotactile feedback device that mapped body tilt estimates onto their torsos via a 3-row by 16-column array of tactile vibrators (tactors). Four tactor display configurations with spatial resolutions ranging between 22.5 degrees and 90 degrees, in addition to the tactors off configuration, were evaluated. Power spectral density functions of body sway in the anterior-posterior (A/P) and medial-lateral (M/L) directions, and transfer functions between platform motion and body sway, were computed at frequencies ranging from 0.0178 to 3.56 Hz. Cross-spectral analysis revealed that the A/P responses were not significantly driven by M/L inputs, and vice versa, thus supporting the notion of independent A/P and M/L postural control. Vibrotactile feedback significantly decreased A/P and M/L spectral power, decreased transfer function gains up to a frequency of 1.8 and 0.6 Hz in the A/P and M/L directions, respectively, and increased phase leads above 0.3 Hz. SDA showed significantly decreased transition time for both A/P and M/L tilts, and decreased transition displacement and short-term diffusion coefficients for A/P tilt. However, the spatial resolution of the tactor displays did not affect subjects' performance, thereby supporting the use of a lower spatial resolution display in future device designs.