We used vertical seismic profiling (VSP) data collected in four shallow boreholes (about 40 to 60 m deep) to study the shear-wave attenuation in the Mississippi embayment in southwestern Tennessee. The source was an air-powered hammer that produces repeatable SH waves that were recorded by monitor geophones deployed on the surface very close to the source. The spectral ratio method was used to estimate the shear-wave quality factor (QS). The method assumes that the amplitudes of the seismic waves decay exponentially in the frequency domain. The spectral ratio was computed using a VSP trace at a certain depth and the corresponding monitor trace. Using the source monitor trace as reference eliminates possible artifacts that may be introduced by changes in the source-ground coupling. The slope of a least-square line fitted to the logarithm of the spectral ratio versus frequency gives the attenuation coefficient α(z) for that depth. Then a straight line is fitted to α(z) as a function of z, and the slope of this line is used to estimate an average value of QS. For the four sites, Shelby Farms, Covington, Brownsville, and Jackson, the QS values are 25.3±2.6, 18.6±3.4, 18.4±7.6, and 16.4±2.6, respectively. To assess the reliability of these results, synthetic VSP data were generated from the first cycles of the monitor traces recorded in the four sites and these QS values. Except for Brownsville, good matches with correlation coefficients equal to 96%–99% were obtained between the synthetic VSPs and the actual VSPs, indicating that the QS values we obtained are reliable. In addition, these values are also consistent with recent results obtained by others using seismic data generated by small earthquakes recorded by the Japanese strong-motion network Kiban-Kyoshin (KiK-net).