Unambiguous identification of polyoxometalate (POM) species generated in self-assembly reactions in solution is rather problematic due to close similarity of their properties such as solubility and spectral characteristics. The situation is made more complex by protonation equilibria (which can change their analytical signals) and the lack of individual compounds to serve as standards for individual members of these mixtures. In the present work a new approach to the study of such POMs has been suggested, taking molybdovanadates [PMo12−xVxO40]−3−x as a model. The key feature of this approach consists of generation of so-called “conditional model systems” that include most of the expected components of a mixture formed by self-assembly, tracked down by reliable detection techniques, e.g., 51V NMR-spectroscopy in this particular case. Then the proposed composition of the mixture is verified and corrected by means of high-performance liquid chromatography coupled with inductively coupled plasma atomic emission spectrometry (HPLC-ICP-AES).