We measured the permeability of 30 samples extracted from 6 sets of compaction bands and the adjacent host rocks of the Jurassic aeolian Aztec Sandstone exposed in the Valley of Fire State Park in Nevada using core flooding experiments. The results show that the permeability within the high-angle compaction bands (three sets) is consistently three orders of magnitude lower than that of the host rocks. For the bed-parallel compaction bands, the measured permeability reduction is about half an order to three orders of magnitude for two sets of bands, and there is no detected permeability reduction for the samples from one set. For the samples that show permeability reduction within high-angle and bed-parallel compaction bands, the results are generally consistent with the data estimated from two-dimensional segmented image analyses in previous studies. Permeability of the samples used in the laboratory experiments was also obtained numerically based on three-dimensional tomographic images scanned from micro-samples and lattice-Boltzmann flow simulations. In addition, backscatter electron images (BET) and energy dispersive spectroscopy images (EDSI) of thin sections were used to estimate the clay content inside and outside the bands. Large differences exist between the lab-based and image-based permeability and porosity measurements of compaction bands and host rocks. Possible factors causing these differences are different sample sizes and heterogeneities within the host rocks, calibration on the image segmentation, and incomplete characterization of clay minerals and fines migration during lab-based experiments. Given the wide range of permeability reductions within compaction bands of different orientations by different investigators, their impact on fluid flow should be evaluated case by case, one should consider their dimensions and distributions.