Seismic studies in the near surface typically involve path lengths that are on the order of a wavelength. However, most investigators extrapolate seismic imaging and parameter extraction techniques to the near surface without incorporating the near-field in their processing approach. For such short path lengths, we find that the near-field is strongly present in the recorded waveforms, which consequently biases seismic parameter estimation, such as Q measurement. To suppress the near-field effects, we examined the challenges of near-field removal and the issues involving the accurate modeling of the near-field. Using the mathematical expression for a source monopole, which describes an elastic impact source typically used in shallow seismic studies, we modeled the near- and far-field terms of the waveform with distance in a homogeneous elastic medium. For thiscase, we found that the near-field can be suppressed through frequen-cy-domain or polarization filtering. However, in heterogeneous media, the near-field cannot be removed through filtering due to interference of scattered P- and S-converted waves. Hence, in realistic media it is necessary to compensate for the near-field bias in Q measurement via appropriate numerical modeling in 3D. To assess the near-field bias in real data, a vertical seismic profiling study was conducted at a site in the Lawrence Livermore National Laboratory, and it was found that the apparent attenuation due to the near-field is of order comparable to the combined effects of scattering and intrinsic attenuation. These results raise caution for the interpretation of extracted Q values in the near surface if they do not account for the near-field presence and point to the necessity of incorporating a near-field term in standard seismic processing techniques when applied to the near surface.