To ensure the safety of drinking water with enough residual chlorine, a part of residual chlorine named "ineffective chlorine" was uniquely identified and quantified by NaAsO2 selective quenching method. NaAsO2 showed a weak reducibility and was easier to be controlled compared to other common quenchers including NaHSO3, ascorbic acid, Na2SO3.and Na2S2O3. Due to the different oxidizing abilities among the species in total residual chlorine, NaAsO2 can selectively and completely quench the strong oxidizing residual chlorine parts (free chlorine, NH2Cl, NHCl2 and some organic chloramines) but leave the weak or no oxidizing parts for the quantification as ineffective chlorine. During the chlorination of Suwannee River humic acid, ineffective chlorine formation increased from <0.02 to 0.08 mg/L (as Cl-2) as the dissolved organic carbon of humic acid increased from 0 to 15 mg-C/L. Finished water samples from three drinking water treatment plants (DWTPs) and one tap water sample were collected to examine the presence of ineffective chlorine. The quantified ineffective chlorine ranged from 3.3% to 27.8% of the total residual chlorine. Formation of ineffective chlorine was positively correlated with the dissolved organic nitrogen concentration in drinking water. As the reaction time increased from 4 to 48 h, the ineffective chlorine in the finished water from DWTP1 (water source: Yangtze River) was fairly stable at 0.15 mg/L as Cl-2, and its proportion in total residual chlorine increased from 27.8% to 78.9%. The ineffective chlorine was verified to be much weaker than free chlorine, NH2Cl and NHCl2 in Escherichia coli inactivation. (C) 2015 Elsevier B.V. All rights reserved.