This study aims to evaluate the influence of epistemic uncertainties in shear wave velocity ( VS) on seismic ground response analyses (GRAs). A number of alternative VS profiles obtained from both invasive (i.e., borehole) and noninvasive (i.e., surface waves) testing methods are available for two blind study sites. These profiles are used to estimate epistemic uncertainties in VS, which are then propagated through equivalent linear-elastic GRAs, allowing for the quantification of intramethod uncertainty and intermethod variability in terms of spectral accelerations, amplification functions, and damage parameters (i.e., Arias and Housner intensities). This study demonstrates that nonlinearity of soil response plays a fundamental role in increasing the propagated uncertainty in GRA. Additionally, GRAs were also performed by means of VS upper/lower-range profiles developed from assumptions commonly used to account for epistemic uncertainties. These VS profiles were found to yield unrealistic site response estimates for both low strain (i.e., linear-elastic) and high strain (i.e., nonlinear) levels.