Résumé: Otolith microchemistry can provide valuable information about stock structure and mixing patterns when the magnitude of environmental differences among areas is greater than the cumulative influence of any vital effects. Here, the current understanding of the underlying mechanisms governing element incorporation into the otolith is reviewed. Hard and soft acid and base (HSAB) theory is employed to explore the differences in chemical behaviours, distributions and affinities between elements. Hard acid cations (e.g. Mg2+, Li+ and Ba2+) tend to be less physiologically influenced and accepted more readily into the otolith crystal lattice but are relatively homogeneous in seawater. Soft acid cations (e.g. Zn2+ and Cu2+) on the other hand, exhibit more varied distributions in seawater, but are more likely to be bound to blood proteins and less available for uptake into the otolith. The factors influencing the geographical distribution of elements in the sea, and their incorporation into the otoliths of marine fishes are reviewed. Particular emphasis is placed on examining physiological processes, including gonad development, on the uptake of elements commonly used in population studies, notably Sr. Finally, case studies are presented that either directly or indirectly compare population structuring or movements inferred by otolith elemental fingerprints with the patterns indicated by additional, alternative proxies. The main obstacle currently limiting the application of otolith elemental microchemistry to infer movements of marine fishes appears to lie in the largely homogeneous distribution of those elements most reliably measured in the otolith. Evolving technologies will improve the discriminatory power of otolith chemistry by allowing measurement of spatially explicit, low level elements; however, for the time being, the combination of otolith minor and trace element fingerprints with alternative proxies and stable isotopic ratios can greatly extend the scope of migration studies. Among the otolith elements that routinely occur above instrument detection limits, Ba, Mn and Li were deemed the most likely to prove reliable geographic markers in marine species.