【Objective】Establishing ecology-based soil heavy metal thresholds is critical for accurately assessing regional ecological quality, managing contaminated sites effectively, and promoting the sustainable development of soil ecosystems. This study aims to derive such thresholds for copper (Cu), zinc (Zn), lead (Pb), and cadmium (Cd) under different land use patterns by integrating extensive ecotoxicological data and accounting for key soil properties. 【Method】It was systematically compiled and screened ecotoxicological data, yielding 118, 123, 61, and 100 valid entries for Cu, Zn, Pb, and Cd, respectively. The dataset covered 29 terrestrial plant species, 4 terrestrial invertebrate species, and 10 soil ecological process indicators, with accompanying soil properties including pH, clay content (CL), organic matter (OM), and cation exchange capacity (CEC). The data were normalized, accounting for aging-leaching effects and soil property influences. Predictive models for 10% effect concentration (EC10) were developed based on soil properties. Species sensitivity distribution (SSD) was applied using weighted average models derived from multiple best-fitting distribution functions to reduce uncertainty related to reliance on a single distribution. 【Result】The result revealed that (1) Soil pH was the primary factor influencing the ecotoxicity of the heavy metals, with significantly higher thresholds observed in alkaline soils compared to acidic soils. (2) The order of ecotoxicity was Cd >> Cu > Zn > Pb. (3) Thresholds varied substantially across land use types: agricultural land and protected natural areas showed the lowest thresholds, reflecting the strictest ecological protection requirements. However, park and residential lands presented intermediate values, while commercial & service and industrial & mining lands had the highest thresholds. (4) The derived ecological thresholds based on land use and pH-specific categories differed notably from current standards in Europe and the United States, mainly due to differences in derivation methods, ecological receptor coverage, protection levels, and incorporation of soil properties. 【Conclusion】This study compiled extensive ecotoxicological data for Cu, Zn, Pb, and Cd and developed toxicity threshold prediction models incorporating key soil properties (pH, CL, OM, CEC) following aging-leaching normalization and soil property normalization. The results provide a scientifically supported and practical framework for refining soil environmental quality standards and support differentiated ecological risk management based on land use and soil properties.