It generally is unknown to what extent the toxicity of Cu and Ni for soil microbial processes varies among different soils. A comparative study was made using three different microbial assays (nitrification potential, glucose-induced respiration, and C-mineralization of a plant residue) in 19 (for Cu) or 16 (for Ni) soils with contrasting soil properties. Each soil was spiked with CuCl2 or NiCl2 at seven different concentrations, and the bioassays were started after a 7-d equilibration period. The Cu and Ni toxicity thresholds varied 19- to 90-fold among soils. The differences in both Cu and Ni toxicity among soils were neither explained by soil solution metal concentrations nor by free ion activities calculated from soil solution composition. Copper toxicity thresholds (total concentrations) increased with increasing organic matter content or cation exchange capacity (CEC) and, surprisingly, decreased with increasing pH depending on the assay. Nickel toxicity thresholds consistently increased with increasing CEC, background Ni, and clay content for all three assays. Thresholds expressed as soil solution free ion activities all significantly decreased with increasing soil solution pH (r2 = 0.57-0.93), consistent with a decreased H+:M2+ competition at the biological membrane. That competition largely counteracts the H+:M2+ competition for sorption, effectively explaining the insignificant or weak effect of pH on total Ni or Cu toxicity thresholds. It is concluded that free metal-ion activity alone does not explain variation in metal toxicity among soils.