The enrichment of soil arsenic (As) and antimony (Sb) is putting increasing pressure on the environment and human health. The biogeochemical behaviour of Sb and its uptake mechanisms by plants are poorly understood and generally assumed to be similar to that of As. In this study, the lability of As and Sb under agricultural conditions in historically contaminated soils was assessed. Soils were prepared by mixing historically As and Sb-contaminated soil with an uncontaminated soil at different ratios. The lability of As and Sb in the soils was assessed using various approaches: the diffusive gradients in thin films technique (DGT) (as CDGT), soil solution analysis, and sequential extraction procedure (SEP). Lability was compared to the bioaccumulation of As and Sb by various compartments of radish (Raphanus sativus) grown in these soils in a pot experiment. Irrespective of the method, all of the labile fractions showed that both As and Sb were firmly bound to the solid phases, and that Sb was less mobile than As, although total soil Sb concentrations were higher than total soil As. The bioassay demonstrated low bioaccumulation of As and Sb into R. sativus due to their low lability of As and Sb in soils and that there are likely to be differences in their mechanisms of uptake. As accumulated in R. sativus roots was much higher (2.5–21 times) than that of Sb, while the Sb translocated from roots to shoots was approximately 2.5 times higher than that of As. As and Sb in R. sativus tissues were strongly correlated with their labile concentrations measured by DGT, soil solution, and SEP. These techniques are useful measures for predicting bioavailable As and Sb in the historically contaminated soil to R. sativus. This is the first study to demonstrate the suitability of DGT to measure labile Sb in soils.