Abstract The object of this study was to investigate the synthesis of a nanoscale zero-valent iron slurry (NZVIS) for use in Fenton-like reactions, and to evaluate its efficiency for As(III) oxidation to As(V) in spiked deionized water and simulated groundwater containing humic acid. Furthermore, this study used injection of the nanoiron slurry combined with electrokinetic processes to remediate As(III) in soil. NZVI was prepared by a chemical reduction process. The efficiency of using 3 wt% soluble starch (SS) to stabilize NZVI was also studied. It was found that the SS could keep the nanoparticles dispersed for over one day. The NZVI was characterized by XRD, FE-SEM, ESEM-EDS, and EDS-mapping, to observe its morphology and crystal structure. In this research the iron species observed took non-crystalline forms. In water batch tests, studies in deionized water were compared with those in simulated groundwater with humic acid, and dissolved oxygen content was adjusted. Injection of NZVIS oxidized As(III) to As(V) in all cases. In both deionized water and simulated groundwater, it was found that when the dissolved oxygen(DO) content was not increased, the NZVIS generated non-selective oxidant OH¡E, thus reducing the As(V) production rate. When dissolved oxygen content was increased, the DO oxidized organic matter present in the simulated groundwater, allowing the OH¡E to react further with As(III) and increasing the As(V) production rate. Finally, a test was performed in actual groundwater under optimal reaction conditions, without increasing the dissolved oxygen content, for comparison of As(V) yield. The concentration of As(V) was found to be higher in this test (As(V) Conc. = 17.55 £gg/L) than when using simulated groundwater (As(V) Conc. = 4.63 £gg/L). This study further examined NZVIS injection combined with electrokinetic (EK) technology for the remediation of soil columns containing a low concentration (initial conc. = 100 mg/kg) and a high concentration (initial conc. = 500 mg/kg) of As(III). EK alone without injection of NZVIS (Test E-1) resulted in a residual soil As(V) concentration of 24 mg/kg in the low-concentration test group. In Test E-2, where NZVIS was injected into the anode reservoir, and Test E-3, where NZVIS was injected into the cathode reservoir, residual soil As(V) concentrations were 2.3 mg/kg and 3.4 mg/kg, respectively. The high-concentration test group was comprised of Test E-4 (EK alone without injection of NZVIS), Test E-5 (NZVIS injected into anode reservoir), and Test E-6 (NZVIS injected into cathode reservoir). In these tests, only soil sections 0.2 and 0.4 (normalized distance from anode reservoir) met soil regulation standards. Residual As(V) concentrations in soil sections 0.6, 0.8, and 1.0 are much higher than the regulatory standard. In soil section 1.0, the residual As(V) concentration was less in Test E-6 than in Test E-5 (116.6 mg/kg and 183.5 mg/kg, respectively). This may be because at high pH values, the iron surface does not corrode, instead arsenic adsorption prevails. Only a fraction of negatively charged As(V) species will migrate towards the anode resulting in a relatively low soil As(V) concentration near the cathode.