The N, N′-diethyl- p-phenylenediamine (DPD) standard method for the determination of free chlorine by spectroscopic detection at 515 nm has a limited analytical range, 0.7–56 μmol l −1. It is shown that detection performed at 324 nm enhances the method sensitivity by a factor of 2. For analyte concentrations above 100 μmol l −1, a decrease in the absorbance is observed at 515 nm, making the method ambiguous, while the absorbance still increases at 324 nm. By applying partial least squares (PLS) regression based on multiple absorbance values in the interval 200–600 nm, the useful analytical range can be extended further, up to 200 μmol l −1. This improvement is achieved whilst maintaining root mean square errors of cross-validation (RMSECV) of the same order as for the univariate standard method. The multivariate approach also enables identification of samples containing interfering metals, here as Fe 3+, which form light absorbing chelates with EDTA.