Metal sensitivity and toxicity to plants are influenced by not only the concentration and the toxicant types, but also by the life-stage or biological process. In comparison seed germination and the early seedling growth are more sensitive to metal pollution because some of the defense mechanisms have not developed. In this study, early seedling growth of Brassica juncea, Brassica napus, and Raphanus sativus was monitored over 7 days to evaluate physiological responses. 8 soils from abandoned mine sites in South Korea, were collected. Total concentration of heavy metal was measured by Aqua Regia. Water extraction, Toxicity Characterizing Leaching Procedure (TCLP) and Wenzel’s sequential extraction were employed to evaluate extractability. Heavy metal contents in plants were measured by HNO₃ digestion to calculate Biological adsorption Coefficient (BAC). Germination percentage (GP), shoot/root elongation (SE/RE), vegetation index (VI) and water content (WC) were measured as physiological responses (PRs). Pearson’s correlation coefficients were calculated between soil metal contents and plant metal contents. Multiple regression analysis was employed using backward step-wise method for plant metal contents and PRs. As a result, As, Cd, Pb, and Zn were measured as abundant metals in the soils, and their concentrations were 16 -37,190, 2 - 74, 56 - 5,820, and 143-3,495 mg/kg respectively. As and Pb, Cd, and Zn were estimated as highest BAC in B. napus, B. juncea, and R. sativus respectively. In correlation analysis, heavy metal contents of B. napus poorly correlated with heavy metal concentration in all extraction methods, whereas radish was significantly correlated. Strong relationships were found between Heavy metal extracted using the TCLP and heavy metal contents in B. juncea. Heavy metal contents of B. napus were correlated with total heavy metal concentration of aqua regia extraction in comparison with other extractions. All extraction methods were positively correlated in R. sativus, especially, F1+F2 fraction of Wenzel’s sequential extraction showed the best correlation among the extraction. In multiple regression analysis, best modeling among the PRs was estimated in root elongation of B. juncea, and shoot elongation of B. napus and R. sativus. It is likely that (i) B. juncea, B. napus, and R. sativus may have a potential as accumulator of As and Pb, Cd, and Zn respectively. (ii) TCLP and Aqua regia well estimated phyto-available fraction of soils for B. juncea and B. napus, while R. sativus were best described in F1+F2 fraction of Wenzel’s sequential extraction. (iii) Elongation of shoot and root well represented physiological damages in seedling growth. Further efforts are required to consider physical and chemical properties of soils, and their effects on phyto-availability should be employed.