Abstract: Paraquat (PQ) and diquat (DQ) are widely used as non-selective contact herbicides. Several cases involving accidents, suicide, and homicide by PQ or DQ poisoning have been reported. Poising by PQ, which is mainly concentrated in the lungs, causes acute respiratory distress syndrome and leads to multiple organ toxicity. The toxic effects of DQ are similar to those of PQ but relatively less intense. The mortality rates in PQ and DQ poisoning are high. Simultaneous monitoring of the PQ and DQ concentrations in plasma and urine can provide valuable information for early clinical diagnosis and prognosis. High performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) is the main analytical method used to detect PQ and DQ in plasma and urine. As both these compounds are highly polar and water soluble, they cannot be retained effectively on a reversed-phase column with conventional mobile phases. The separation of PQ and DQ by ion-pair chromatography or hydrophilic chromatography has been reported. The use of an ion-pairing reagent helps in improving the retention capabilities of PQ and DQ. However, the sensitivity of MS detection is noticeably decreased because of ion suppression caused by the ion-pairing reagent in the mobile phase; furthermore, ion-pairing reagents may contaminate the MS system. The separation of PQ and DQ by hydrophilic chromatography is easily affected by matrix components in the sample, and their retention times are not stable. Considering PQ and DQ are bicharged cation species in solution, they are more suitable for separation by cation-exchange chromatography. A method based on ion chromatography-triple quadrupole mass spectrometry was established for the determination of PQ and DQ in plasma and urine. The plasma and urine samples were diluted with water, and then purified on a solid-phase extraction column containing a polymer-reversed phase and weak ion-exchange mixed-mode adsorbent (Oasis WCX). PQ and DQ were separated on an IonPac CS 18 analytical column (250 mm×2.0 mm, 6.0 μm) with gradient elution using a methylsulfonic acid solution electrolytically generated from an on-line eluent generation cartridge. An in-line suppressor was used to remove methylsulfonate and other anions from the eluent before the eluent entered the mass spectrometer. Between the suppressor and the ion source in MS, the addition of 3% (v/v) formic acid in acetonitrile as an organic modifier (using an auxiliary pump and a T-piece) aided desolvation in the ion source, resulted in a one-or two-fold improvement of the response, and eliminated the residual effects of the adsorption of PQ and DQ caused by ion source. The analytes were detected by triple quadrupole tandem mass spectrometry using positive electrospray ionization in the multiple reaction monitoring (MRM) mode. PQ-d₈ and DQ-d₄ were used as internal standards. The calibration curves for PQ and DQ showed good linear relationships in the ranges of 1.0-150 μg/L and 0.5-75 μg/L, respectively, and the correlation coefficients were > 0.999. The average matrix effects of PQ and DQ in plasma were 84.2%-89.3% and 84.7%-91.1%, while the average matrix effects of PQ and DQ in urine were 50.3%-58.4% and 51.9%-59.4%. The average recoveries of PQ and DQ in plasma were 93.5%-117% and 91.7%-112%, respectively, with relative standard deviations (RSDs) of 3.4-16.7% and 2.8%-13.2%, and that in urine were 90.0%-118% and 99.2%-116%, with relative standard deviations of 5.6%-14.9% and 2.4%-17.3% (n=6). The limits of detection of PQ and DQ in plasma and urine were 0.3 μg/L and 0.2 μg/L, respectively, with the corresponding limits of quantification being 1.0 μg/L and 0.5 μg/L. This method is sensitive and accurate, and it can be used to determine PQ and DQ for clinical diagnosis and prognosis in patients.