Spin-echo fMRI offers a potentially better intrinsic functional spatial resolution than its gradient echo counterpart, as well as the elimination of signal dropouts in the image. This comes at the price of a significant loss in sensitivity. In this article the main methods for measuring spin-echo fMRI are presented: HASTE, SSFP, RASER and most importantly spin-echo EPI. Their relative merits and limitations are discussed. The BOLD contrast mechanisms responsible for spin echo fMRI are summarised, and the spatial origin of the signal within the neocortex discussed. The major publications concerning the use of spin echo fMRI are examined. At present the most promising application for this methodology would appear to be in the examination of cortical layers and columns. The balance of experimental and theoretical evidence accumulated to date leads the author to propose that: (i) There is little point in conducting spin-echo fMRI at main magnetic field strengths of 3 T and below; (ii) There are fundamental limitations to acquiring spin-echo BOLD data at 7 T and above; (iii) Whole brain coverage with SE-BOLD at very high static magnetic field strengths could prove valuable; and (iv) SE-BOLD is probably better suited to study cortical columns than cortical layers. Recently gradient-echo approaches for high spatial resolution fMRI have been demonstrated that employ special techniques to avoid the effects of larger post capillary vessels. The coming years will show whether spin-echo techniques can remain the method of choice for high spatial resolution studies, and whether they can extend their range of application at 7 T and above.