Extremely low limits of detection in trace element analysis can be achieved when coupling generation of volatile hydrides to the relatively simple methods of analytical atomic spectrometry - atomic fluorescence spectrometry (AFS) or atomic absorption spectrometry (AAS). To reach ideal performance of the whole analytical procedure, the atomizers dedicated for coupling with AAS or AFS (conventional externally heated quartz tubes - CQTA, miniature diffusion flames - MDF and dielectric barrier discharges - DBD) have to be optimized in terms of design as well as of operation parameters. Such an optimization can be made in a straightforward and elegant way based on the knowledge of what really happens in hydride atomizers. The key point which must be taken into account in order to understand what really happens in these atomizers is that their temperature is too low to be compatible with any significant thermal atomization of hydrides. The dramatic disagreement with the many years of experience of observing a complete conversion of hydrides to free atoms is explained by the radical theory of hydride atomization. The presented evidence corroborates the radical theory of hydride atomization in the CQTA. This makes possible optimization of design as well as of operational parameters of this kind of atomizer just on the basis of quantification of distributions of hydrogen radicals which can be determined either experimentally by two-photon absorption laser-induced fluorescence or potentially by numerical simulation. Regarding extension of the radical theory in the CQTA to MDF and DBD atomizers, more experimental evidence on free analyte atom distributions is required either to confirm its validity or to discover reasons for its failure.