Arsenic is a widespread contaminant in soils and groundwater. While various iron-based materials have been studied for immobilizing arsenic in contaminated soils, the feasibility of stabilized iron-based nanoparticles has not been reported. This study investigates the effectiveness of using three types of starch-stabilized iron-based nanoparticles, including zero-valent iron (ZVI), iron sulfide (FeS), and magnetite (Fe3O4), for immobilization of arsenic in two representative As-contaminated soils (an orchard soil and a fire range soil). To test the effect of the nanoparticles on the arsenic leachability, As-contaminated soils were amended with the nanoparticles at various Fe/As molar ratios (5:1―100:1) and contact time (3 and 7 d). After three days’ treatments of a field-contaminated sandy soil, the PBET-based bioaccessibility of As decreased from an initial (71.3±3.1)% (mean±SD) to (30.9±3.2)% with ZVI, (37.6±1.2)% with FeS, and (29.8± 3.1)% with Fe3O4 at an Fe/As molar ratio of 100:1. The TCLP-based leachability of arsenic in a spiked fire range soil decreased from an initial (0.51±0.11)% to (0.24±0.03)%, (0.27±0.04)% and (0.17±0.04)% by ZVI, FeS, and Fe3O4 nanoparticles, respectively. The Fe3O4 nanoparticles appeared to be more effective (5% or more) than other nanoparticles for immobilizing arsenic. When the two soils were compared, the treatment is more effective on the orchard soil that has a lower iron content and higher initial leachability than on the range soil that already has a high iron content. These results suggest that these innocuous iron-based nanoparticles may serve as effective media for immobilization of As in iron-deficient soils, sediments or solid wastes. Arsenic is a widespread contaminant in soils and groundwater. While various iron-based materials have been studied for immobilizing arsenic in contaminated soils, the feasibility of stabilized iron-based nanoparticles has not been reported. This study investigates the effectiveness of using three types of starch -stabilized iron-based nanoparticles, including zero-valent iron (ZVI), iron sulfide (FeS), and magnetite (Fe3O4), for immobilization of arsenic in two representative As-contaminated soils (an orchard soil and a fire range soil). To test the effect of the nanoparticles on the arsenic leachability, As-contaminated soils were amended with the nanoparticles at various Fe / As molar ratios (5: 1-100: 1) and contact time (3 and 7 d). After three days ’treatments of a field-contaminated sandy soil, the PBET-based bioaccessibility of As decreased from an initial (71.3 ± 3.1)% mean ± SD to 30.9 ± 3.2% with ZVI, (37.6 ± 1.2)% with FeS , and (29.8 ± 3.1)% with Fe3O4 at an Fe / As molar ratio o f 100: 1. The TCLP-based leachability of arsenic in a spiked fire range soil decreased from an initial (0.51 ± 0.11)% to (0.24 ± 0.03)%, (0.27 ± 0.04)% and (0.17 ± 0.04)% by The Fe3O4 nanoparticles were more effective (5% or more) than other nanoparticles for immobilizing arsenic. When the two soils were compared, the treatment is more effective on the orchard soil that has a these results suggest that these innocuous iron-based nanoparticles may serve as effective media for immobilization of As in iron-deficient soils, sediments or solid wastes.