The spectral-analysis-of-surface-waves (SASW) method is a relatively new in situ method for determining shear-wave slownesses. All measurements are made on the ground surface, making it much less costly than methods that require boreholes. The SASW method uses a number of active sources (ranging from a commercial Vibroseis truck to a small handheld hammer for the study conducted here) and different receiver spacings to map a curve of apparent phase velocity versus frequency. With the simplifying assumption that the phase velocities correspond to fundamental mode surface waves, forward modeling yields an estimate of the sub-surface shear-wave slownesses. To establish the reliability of this indirect technique, we conducted a blind evaluation of the SASW method. SASW testing was performed at 10 strong-motion stations at which borehole seismic measurements were previously or subsequently made; if previously made, the borehole results were not used for the interpretation of the SASW data, and vice-versa. Comparisons of the shear-wave slownesses from the SASW and borehole measurements are generally very good. The differences in predicted ground-motion amplifications are less than about 15% for most frequencies. In addition, both methods gave the same NEHRP site classification for seven of the sites. For the other three sites the average velocities from the downhole measurements were only 5-13 m/sec larger than the velocity defining the class C/D boundary. This study demonstrates that in many situations the SASW method can provide subsurface information suitable for site response predictions.