A proposed method for dealing with foreground emission in upcoming 21-cm observations from the epoch of reionization is to limit observations to an uncontaminated window in Fourier space. Foreground emission can be avoided in this way, since it is limited to a wedge-shaped region in k∥, k⊥ space. However, the power spectrum is anisotropic owing to redshift-space distortions from peculiar velocities. Consequently, the 21-cm power spectrum measured in the foreground avoidance window – which samples only a limited range of angles close to the line-of-sight direction – differs from the full redshift-space spherically averaged power spectrum which requires an average over all angles. In this paper, we calculate the magnitude of this ‘wedge bias’ for the first time. We find that the bias amplifies the difference between the real-space and redshift-space power spectra. The bias is strongest at high redshifts, where measurements using foreground avoidance will overestimate the redshift-space power spectrum by around 100 per cent, possibly obscuring the distinctive rise and fall signature that is anticipated for the spherically averaged 21-cm power spectrum. In the later stages of reionization, the bias becomes negative, and smaller in magnitude (≲20 per cent).