Toxicity of metals released from mine tailings may cause severe damage to ecosystems. A diversity of microorganisms, however, have successfully adapted to such sites. In this study, our objective was to advance the understanding of the indigenous microbial communities of mining-impacted soils. To this end, a metatranscriptomic approach was used to study a heavily metal-contaminated site along a metal concentration gradient (up to 3220 000 and 97 000 mg kg⁻¹ of Cd, Pb and Zn, respectively) resulting from previous mining. Metal concentration, soil pH and amount of clay were the most important factors determining the structure of soil microbial communities. Interestingly, evenness of the microbial communities, but not its richness, increased with contamination level. Taxa with high metabolic plasticity like Ktedonobacteria and Chloroflexi were found with higher relative abundance in more contaminated samples. However, several taxa belonging to the phyla Actinobacteria and Acidobacteria followed opposite trends in relation to metal pollution. Besides, functional transcripts related to transposition or transfer of genetic material and membrane transport, potentially involved in metal resistance mechanisms, had a higher expression in more contaminated samples. Our results provide an insight into microbial communities in long-term metal-contaminated environments and how they contrast to nearby sites with lower contamination.