After a literature survey and observations made during a space experiment, the phenomenon of crucible de-wetting by the crystal during Bridgman solidification in microgravity is explained by a model involving composite wetting of the crucible by the liquid, crystal angle of growth and interface advance. A ground experiment was run in order to validate this model which also explains why a crystal detaches from the crucible surface when a sand blasted crucible is used in Bridgman solidification on the ground. It is shown that de-wetting leads to enhanced quality of the crystal produced and that capillary-induced convection effects are not to be feared in this case. Consequently, it is highly advisable to use rough-surface crucibles for crystal growth both in microgravity and on the ground. © 1990.