It is known from the classical nucleation theory that crystallization can only occur if critical nuclei are formed. Thus, there has to be enough space available for the formation of such nuclei. In the present work, we determined theoretical criteria that have to be satisfied for the occurence of a crystalline phase from a supersaturated solution in confinement conditions (e. g. resulting from entrapment of a drug solution into a porous matrix). Similar criteria were determined for the occurence of crystallization from amorphous or metastable polymorphic phases in confinement conditions. It was shown theoretically that spatial constraint can cause vitrification from solution or polymorph selection1. Metastable phases, which are formed in this manner, can be effectively protected against crystallization into thermodynamically more stable products through space restriction. The theoretical criteria were tested experimentally with nifedipine entrapped into a nanoporous silica xerogel matrix with an average pore diameter of 2,5 nm. The results of thermal analysis and X-ray powder diffraction have shown that the amorphous nifedipine, formed inside the xerogel pores, is protected against crystallization in the temperature interval in which it is chemically stable. With a combination of differential scanning calorymetry and BET analysis we estimated the total amount of nifedipine that can be stabilized in the xerogel. Also, it was shown through calculations, that entrapment into a hypotetical nanoporous matrix with a certain pore size allows selective formation of sulfathiazole polymorphs. If the pore radius is equal to 1,7 nm, then form IV is formed selectively from a toluene solution. If the pore radius is less than 1,4 nm, an amorphous phase is formed.