Copper toxicity in 17 Chinese soils was screened using a substrate-induced nitrification assay to generate information for the development of a terrestrial biotic ligand model (tBLM). The leaching effect on the Cu toxicity thresholds was investigated. Both the total Cu-based median effective concentration (EC50) values (46.9–2726 mg/kg) and the solution Cu-based EC50 values (0.04–2.91 mg/L) in unleached soils varied substantially among the soils in the present study. For unleached soils, linear stepwise multiple regression analysis showed that total Ca and soil pH were the best predictors for total Cu-based EC50, while electrical conductivity (EC) and soil pH were the best predictors for solution Cu-based EC50. The variation in solution Cu-based EC50 was largely (R2 1/4 0.75) explained by Mg but not Ca and Ht concentration in soil solution at EC50, suggesting a protective effect of Mg2t against Cu toxicity in the test soils. Leaching impacted Cu toxicity differently among the soils and apparently reduced the variations of both the total Cu-based and the solution Cu-based EC50. The predictability of the Cu EC50 by empirical models was decreased after leaching. The leaching effect on Cu toxicity, indicated by a leaching factor, was not predicted by any soil properties. There is a need to investigate quantitatively the mechanisms for the leaching effect on Cu toxicity in soils.