AbstractToxic effects of heavy metals on soil microorganisms have been confirmed in a number of laboratory studies. However, most real‐field studies do not allow for strong general conclusions due to a range of problems, such as pseudoreplication and confounding factors, which are almost impossible to control for with the most commonly used polluted versus unpolluted or random sampling designs. Effects of metal contamination on soil microbial community traits were measured along 2 pollution gradients in southern Poland. Employing an experimental regression design, using 2 separate gradients, the authors aimed to control for effects of soil properties and beta‐diversity of microbial communities. General microbial activity was measured as soil basal respiration rate and substrate‐induced respiration, while microbial functional and structural diversity were analyzed with community‐level physiological profiles and phospholipid fatty acid patterns, respectively. Metal concentrations were normalized to their toxicity and integrated in a toxicity index (TI). Microbial activity (basal and substrate‐induced respiration) decreased in both gradients with increasing TI. Community‐level physiological profiles for fungi correlated positively with TI, but no impact of TI on the community‐level physiological profiles of bacteria was observed. The phospholipid fatty acids a:15 and i:17 were positively correlated with TI, whereas 16:1ω9 and 18:2ω9 were negatively correlated with TI. The use of 2 gradients (Olkusz and Miasteczko Śląskie) allowed the authors to reveal a clear effect of pollution on general microbial structure and activities, even though they were not able to control completely for all confounding factors. Soil pH, organic matter content, and nutrient level appeared to be at least as important as TI in determining microbial community structure and activities. Environ Toxicol Chem 2013;32:1992–2002. © 2013 SETAC