Hydrolysis of metal alkoxides or metal salts often results in the formation of charge- or ligand-stabilized molecular clusters (polynuclear metal oxo compounds) instead of gels or microcrystalline precipitates. In most cases, reliable protocols are at hand to synthesize either clusters or gels deliberately. Nevertheless, a prediction of the type of compound formed from a particular precursor under a certain set of reaction conditions is not possible. Many other issues are also unresolved. For example: Are clusters “dead ends” or is their formation reversible? Are clusters building blocks of gels? Why are many large clusters odd-shaped (unlike nanoparticles)? Are clusters structural models of oxides? How are large clusters formed? The talk will summarize, for both silicon and transition metal compounds, what is known and what is not known, and will address research deficits.