Large-scale groundwater contamination by arsenic (As) requires economical yet reliable analytical procedures to monitor the quality of well water and the effluent of As elimination processes. The silver diethyldithiocarbamate (AgDDTC) method provides the basis of a series of colourimetric tests used as a compromise between semi-quantitative field-test kits and expensive instrumental analytics. This study aimed at understanding the significant As overestimation by the AgDDTC method observed in landfill leachate samples. Results of the AgDDTC test were up to threefold the levels obtained by reference analyses. Overestimation broadly agreed with the time course of chemical oxygen demand and total phosphorus in the leachate. We hypothesised that the interference was caused by reduced P species or humic acid (HA), respectively. To test this, As-AgDDTC determinations from As standards were conducted with phosphonic (H 3 PO 3 ) and phosphinic acid (H 3 PO 2 ) as well as two commercial HA as matrix components. While H 3 PO 3 showed no effect, H 3 PO 2 in concentrations of 0.06 to 28 mg P L −1 yielded As recoveries of up to 130%. This was probably due to the evolution of phosphine (PH 3 ) as suggested by the absorption spectra recorded after reacting pure PH 3 with AgDDTC in pyridine. The HA in concentrations of 0.6-320 mg dissolved organic carbon L −1 gave As recoveries of up to 600%, possibly due to a chromophoric reaction of volatile HA digestion products formed during the reduction step. Significant As overestimation in the presence of the above-named compounds may require pre-oxidation of the samples. To recognise interference effects, the acquisition of full absorbance spectra may be recommendable in lieu of fixed wavelength analyses.