Abstract Trace metal concentrations were measured in the Danube River at different locations along the three main branches of the delta. Results from a systematic delta survey in 1997 and those from the 1995 cruise indicate that the ranges of the total dissolved concentrations in the river end-member are: A1=211–445 nM, Cd=224–302 pM, Co=266–390 pM, Cu=61–120 nM, Fe=21–156 nM, Mn=26–55 nM, Pb=145–220 pM and Zn=11–41 nM. The particulate concentrations were found to be: A1=5·5–6·7%, Cd=1·1–2·4 ppm, Co=16–19 ppm, Cu=201–1092 ppm, Fe=3·7–4·0%, Mn=1286–2290 ppm, Ni=66–77 ppm, Pb=58–65 ppm and Zn=212–224 ppm. The concentrations measured in 1997 confirm that the total dissolved concentrations in the Danube River are low and do not give any evidence of contamination, except for Cu. The particulate concentrations of A1, Ni, Cd, Co and Mn are very similar in the Danube and in uncontaminated world rivers. Pb is found to be enriched by a factor of 3 and Cu by a factor of 14. This enrichment may be due to some mining activity occurring upstream and within the Danube delta area. The evolution of the total dissolved metal concentration (Cd, Cu, Fe, Mn and Zn) along the salinity gradient established in 1997 for the Chilia and Sulina branches confirmed the low reactivity of dissolved trace metals. The only exception being an addition of manganese in front of the Chilia branch. This excess may be attributed to the occurrence of highly labile particulate manganese originating from a local source within the Chilia branch, probably from mining waste. The comparison between field data for winter and spring conditions do not show any significant difference for Cd, Mn and Zn. The comparison with results from the in vitro mixing of river and seawater seems to indicate that the processes involved in the addition of Cd and Zn are inorganic. In winter, the total dissolved iron concentrations were found to be pH-dependent, whereas in spring, due to the beginning of photosynthetic activity, the pH increased within the estuary, and the lower concentrations of Fe observed (down to around 10nM) can be considered as an indirect consequence of the biological activity. The Cu distribution was found to be conservative in winter; in spring, the river concentrations are reduced by a factor of 1·4, which could have resulted from biological uptake by freshwater phytoplankton species. The distribution also seemed to show a slight removal of Cu, similar to the one observed for silica, suggesting that some uptake took place in the low salinity area.