Fibrous sepiolite crystals derive much of their commercial value from their molecular size channels and grooves. The crystals fold upon drying and these channels and grooves are lost. A model for the folding and unfolding of the crystals is presented. Extensive i.r., X-ray and thermogravi- metric evidence shows that folding occurs when approximately half of the water of hydration, which is coordinated to the edge magnesium atoms inside of the channels, is removed. This occurs near 175~ under vacuum and near 300~ in air. When the crystals fold, all remaining water molecules enter a new environment, that of the hexagonal holes of the neighboring silica surface. A true anhydride is produced at about 500~ under vacuum when the final water is lost, but this final dehydration produces no important structural change. Rehydration of the anhydride to the normal hydrated sepiolite does not occur at room temperatures in 100% r.h. However, above, 60~ rehydration does occur. not after the high temperature treatment. Nagata et al. (1974) consider that the folded stage produced at half dehydration will rehydrate, but the true anhy- dride does not rehydrate even under hydrothermal conditions, Our study is designed to clarify the conditions un- der which sepiolite folds and unfolds and the involve- ment of the water of coordination in the folding. In- terpretation of TGA-DTGA. X-ray powder diffrac- tion patterns and, especially, careful study of the major changes observable in the i.r. spectra of the struc- tural OH during the dehydration and folding process will be used to explain a folding model.