Crystallographic texture is generally measured using X-ray diffraction, performed off-line using small samples determining near-surface texture only; electron backscattered diffraction (EBSD) can also be used, but only samples relatively small areas. Ultrasonic methods determine elastic property anisotropy and texture, via orientation distribution coefficients (ODCs), and while there is substantial literature comparing ultrasonically determined properties with X-ray or neutron diffraction texture, there is little discussion about texture inhomogeneity (place to place in a sheet or through thickness) and sampling volume effects (X-ray compared to EBSD) on the accuracy of the correlations. In this article, the crystallographic texture of nominally pure aluminum and commercial aluminum alloy sheets has been determined by X-ray diffraction and EBSD and used to calculate the elastic anisotropy, which is then compared to ultrasonic electromagnetic acoustic transducer (EMAT) velocity anisotropy taking into account through-thickness texture variations. Significant and consistent spatial variability in texture occurs in the aluminum sheet samples (sheet edge to center and through thickness). Predictions of elastic anisotropy based on surface texture determination, as characterized by X-ray diffraction or surface EBSD, gave poor correlations with EMAT velocity anisotropy when the sample contained significant through thickness texture variations; however, accounting for this using multiple EBSD scans through thickness gave good correlations.