Otolith based methods have the potential to discriminate between stocks, an important requirement for sustainable management of fish. The abilities of two otolith based methods to investigate stock structure of the sciaenid Argyrosomus japonicus in South Australia were compared: (i) elemental signatures (Sr:Ca, Ba:Ca, and Mg:Ca) from the otolith edge, and (ii) shape characteristics (otolith morphometrics and overall shape) of whole otoliths. Comparison of elemental signatures indicated that Ba:Ca levels were low in the western coast, intermediate in the central coast and high in the eastern coast. Constrained Canonical Analysis of Principal Coordinates (CAP) allocated elemental concentrations of individual otoliths to regions with 100, 100, and 87% success for western, central and eastern coasts respectively. Otolith shape (elliptical Fourier descriptors and morphological indices) supported results from the elemental study with allocation success of 85, 57, and 85% for western, central and eastern coasts respectively. Shape analysis was then used to investigate the origin of individuals caught in marine waters but suspected of being from an aquaculture facility. The two stock discrimination methods were complementary because trace-element analysis of the otolith edge provided very high classification success and gave a snapshot of differences between groups from different geographic areas, while shape analysis indicated that these discrete groups of fish experienced different environmental conditions over a long period of time. Results from this study highlight the importance of multiple methods in stock discrimination and suggest sub-structuring of the stock of A. japonicus in South Australia.