Wavelet encoding is presented and compared to phase encoding. In wavelet encoding a distribution of spins is excited by a slice selective RF pulse; for each repetition time the distribution excited has the profile of a wavelet at different scale and translation. The spin density can be reconstructed with an inverse wavelet transform. Wavelet encoding has three advantages over phase encoding: (1) there is no Gibb's ringing from partial volume effects, (2) the effective repetition time can be 36 times the repetition time for a 256 x 256 image, and (3) motion artifacts are local and dramatically reduced. Using wavelet encoding, a 256 x 256 T2-weighted projection image can be acquired in 33 s.