Abstract: The method is based on the Cu(II)-catalyzed oxidation of o-dianisidine (I) by S2O82- (II) in the flow injection system of Wentzell et al. (J. Autom. Chem., 1989, 11, 227) to give a product with λmax = 450 nm. The sample (e.g., 100 µL) was injected into a stream of 0.5 mM Cu(II) (1.5 mL min-1) and mixed with a stream of 10 mM I (0.5 mL min-1) in aqueous 40% acetone in a 50-cm coil before being mixed with a stream of 0.1 M 3,3-dimethylglutaric acid (0.5 mL min-1) in 1 M NaCl (pH adjusted to 7.0 with NaOH) in a further 50-cm coil. The manifold was constructed from PTFE tubing (i.d. 0.5 mm). Calibration graphs based on the absorbance at λmax were rectilinear for 25 to 750 µM-II; the detection limit was 0.5 µM. In the determination of 25 µM-II the coefficient of variation was 1.5% (n = 10). There was little interference from 2.5 M methanol, 1.0 M ethanol or 0.25 M prop-2-en-1-ol, some interference from 0.25 M prop-2-enyl acetate and 0.5 M ethyl acetate and serious interference from formaldehyde.