Abstract
For the foreseeable future, progress in determining high-resolution structures of membrane proteins will depend on crystallographic techniques. Until recently, x-ray crystallography seemed to be the only promising method. Progress with this technique, however, has not been as rapid as originally hoped because it is difficult to grow large and sufficiently well-ordered three-dimensional crystals. Electron crystallography of two-dimensional crystals is now a viable alternative that is particularly suitable for structural studies of membrane proteins because of their natural propensity to form two-dimensional arrays.