The reduction of Cr(VI) by the nZVI particles seriously depends on not only the nZVI dosages but also the surface characteristics of nZVI particles caused by different syntheses processes. Therefore, to control the nZVI process only based on the nZVI dosage may not reliable. In this study, a nanoscale, zero-valent iron (nZVI), batch process was used to remove hexavalent chromium Cr(VI) from wastewater. For initial Cr(VI) concentrations of 5, 10, and 20 mg/L, Cr(VI) removal efficiencies were as high as 99% when the nZVI doses were around 0.08, 0.25, and 0.55 g/L, respectively. It was noted that the monitoring of pH, ORP, and DO in the reactor presented good correlations with the Cr(VI) removal efficiencies. Therefore, the monitoring of pH, ORP, DO, the initial Cr(VI) concentration were applied to evaluate the Cr(VI) removal efficiencies. The BPN (Back-Propagation Neural) model provided much more precise predicted Cr(VI) removal efficiencies than the regression models, with the correlation coefficients (R-2) of 0.99, 0.99, and 0.97 for contact times of 30, 60, and 90 min, respectively. As a result, the monitoring pH, OPR, and DO have good capabilities to evaluate the Cr(VI) removal efficiencies, which have high potential for use in controlling the nZVI process for the removal of Cr(VI) from wastewater. (C) 2014 Elsevier B.V. All rights reserved.