Metal contamination of soil has become a serious environmental problem worldwide. Many studies have attempted to infer metal ecotoxicity from soil microbial responses. However, much of the data from these studies tends to be inconsistent and difficult to interpret. We hypothesized that microbial response would be a useful indicator of metal toxicity in soils contaminated by copper mining in Chile. Eighty-four topsoils (0–20 cm) were collected from three areas historically contaminated by copper mining (total Cu: 46–1106 mg kg−1, soluble Cu: 0.05–2.3 mg kg−1, pCu2+: 6.3–10, pH: 5.1–7.8, organic matter: 1.1–10%, clay: 0–28%). Based on soil metal concentrations and ecotoxicity thresholds, Cu was expected to be toxic to microorganisms in the studied soils, while the effects of other metals (total Zn: 79-672, As: 1.9–60, Pb: 19-220, Cd: 0.4–5.1 mg kg−1) were expected to be minor. Soil microbial responses (microbial biomass and numbers, nitrogen mineralization and nitrification, and community-level physiological profiles) were also measured. The results showed that the different responses of soil microbes were not correlated with each other. Furthermore, the soil microbial responses were mainly influenced by the physicochemical properties of the soil, not by the metal concentrations in the soil. The effect of copper on the microbial response was either stimulating (positive) or toxic (negative). Of the soil microbial responses measured in this study, only the microbial biomass was useful for calculating dose-response curves. However, the microbial biomass response was not consistent among the different soil copper pools (total copper, soluble copper, and activity of free Cu2+ ions). It is important to emphasize that the thresholds obtained for copper toxicity cannot be adopted in a robust manner because of the different microbial responses in different sampling areas. Thus, in the copper-contaminated soils under study, microbial response was found to be an unreliable indicator of metal toxicity.