We revisit the well-known Lifshitz-Slyozov model for precipitation, from the perspective of detailed balance equilibria and saturation density It is shown that, in this respect, the Lifshitz-Slyozov model behaves very:differently from its discrete counterpart, the Becker-Doring system; in particular it has no saturation density. We propose a modification of the Lifshitz-Slyozov model which has a saturation density, and whose detailed balance equilibria are a continuous analog of those of the Becker-Doring system. Therefore this model seems more suitable for the study of phase transitions. Mathematically, the modified system consists of a parabolic equation coupled to an integral equation. (C) 1998 Elsevier Science Ltd. All rights reserved.