Abstract
Various theoretical and experimental methods for studying nucleation are considered. Since theoretical methods have fundamental limitations, information is gained from experimental research. It is demonstrated that the nucleation theory is entering a new phase of development, which is going to be, in essence, semiempirical. A semiempirical method for constructing nucleation rate surfaces is described. The topology of these surfaces in the limit of the zero nucleation rate can be determined using phase equilibrium diagrams. This opens up possibilities for revising certain approximations employed in experiments and particular axiomatic statements of the classical nucleation theory. Examples of the practical use of nucleation essential for the development of modern technologies are given.