We have created a benchmark of spatial variations in shear wave anisotropy around Mount Ruapehu, New Zealand, against which to measure future temporal changes. Anisotropy in the crust is often assumed to be caused by stress-aligned microcracks, and the polarization of the fast quasi-shear wave (ϕ) is thus interpreted to indicate the direction of maximum horizontal stress, but can also be due to aligned minerals or macroscopic fractures. Changes in seismic anisotropy have been observed following a major eruption in 1995/96 and were attributed to changes in stress from the depressurization of the magmatic system. Three-component broadband seismometers have been deployed to complement the permanent stations that surround Ruapehu, creating a combined network of 34 three-component seismometers. This denser observational network improves the resolution with which spatial variations in seismic anisotropy can be examined. Using an automated shear wave splitting analysis, we examine local earthquakes in 2008. We observe a strong azimuthal dependence of ϕ and so introduce a spatial averaging technique and two-dimensional tomography of recorded delay times. The anisotropy can be divided into regions in which ϕ agrees with stress estimations from focal mechanism inversions, suggesting stress-induced anisotropy, and those in which ϕ is aligned with structural features such as faults, suggesting structural anisotropy. The pattern of anisotropy that is inferred to be stress related cannot be modeled adequately using Coulomb modeling with a dike-like inflation source. We suggest that the stress-induced anisotropy is affected by loading of the volcano and a lithospheric discontinuity.