The ability of the simultaneously extracted metals/acid volatile sulfides (∑SEM/AVS) index to ascertain environmental risk from potentially toxic elements in calcareous saltmarsh soils was tested using structural equation modeling. This technique allows the detection of both direct and indirect relationships among AVS, SEM, and other soil variables, representing results in a graphical view. The dataset included 90 soil samples from 21 different sites belonging to 6 different saltmarshes and featured a wide range of soil chemicophysical properties. Variables included in the a priori model were hydroperiod, pH, soil redox potential, labile organic carbon, carbonates, total iron, and total amount of potentially toxic elements (PTEs). The best optimized model pointed out the main soil properties that affect AVS accumulation and SEM speciation in these soils. Effect plots of AVS and SEM calculated with the partial linear mixed-effects models included in the piecewise structural equation modeling showed a significant and positive influence of pH and carbonates on AVS and a highly significant effect of carbonates and labile organic carbon on SEM. Single SEM components were also considered separately, to define the potential contributions of labile organic carbon or carbonates as alternative binding phases. Simultaneously extracted Cu, Ni, and Zn were preferentially bound to carbonates, followed by labile organic carbon, whereas Pb and Cd were easily bound to labile organic carbon. Environ Toxicol Chem 2019;38:2688–2697.