A hydroponics experiment was conducted to investigate the effects of iron plaque on root surfaces with respect to selenite uptake and translocation within the seedlings of two cultivars of rice (Oryza sativa L. cv Xiushui48 and Bing9652). Different amounts of iron plaque were formed by adding 0, 10, 30, 50, 70 mg Fe l -1 in the nutrient solution. After 24 h of growth, the amount of iron plaque was positively correlated with the Fe 2+ addition to the nutrient solution. These concentrations of Fe, inducing plaque, had no significant effect on the shoot and root growth of rice plants in 50 lg Se l -1 nutrient solution. The amount of Se accumulated in iron plaque was positively correlated to the amount of iron plaque. Increasing iron plaque decreased the selenium concentration in shoots and in roots. At the same time, the translocation of Se from roots to shoots was reduced with increasing amounts of iron plaque. At both the shorter and longer exposure times, the ratio of root-to-shoot selenium was higher than in the controls. More Se stayed in the roots at the longer exposure time than at the shorter time. The concentration of selenium in the xylem sap was sharply decreased with increasing amount of iron plaque on the rice roots. The DCB (dithionite-citrate-bicarbonate)-extracted Se was up to 89.9-91.1% of the total Se when roots with iron plaque (Fe 70) were incubated in 50 lg Se l -1 solution for 30 min. This DCB-extracted Se, however, accounted for only 21.9-28.7% of total Se when roots with iron plaque were incubated in the same solution for 3 days. Se adsorbed in iron plaque can be desorbed by low-molecular-weight organic acids, similar to the desorption of Se from ferrihydrite. These results suggest that iron plaque might act as a 'buffer' for Se in the rhizosphere.