Two different methods for predicting the quality of the water draining from a pyritic tailing are compared; for this, a static test (ABA test) and a kinetic test in large columns were chosen. The different results obtained in the two experimental set-ups show the necessity of being careful in selecting both the adequate predictive method and the conclusions and extrapolations derived from them. The tailing chosen for the weathering tests (previously tested in shake flasks and in small weathering columns) was a pyritic residue produced in a flotation plant of complex polymetallic sulphides (Huelva, Spain). The ABA test was a modification of the conventional ABA test reported in bibliography. The modification consisted in the soft conditions employed in the digestion phase. For column tests, two identical methacrylate columns (150 cm high and 15 cm diameter) were used to study the chemical and microbiological processes controlling the leaching of pyrite. The results obtained in the two tests were very different. The static test predicted a strong potential acidity for the tailing. On the contrary, pH value in the effluents draining from the columns reached values of only 5 units, being the concentration of metals (<600 mg/L) and sulphate ions (<17,000 mg/L) very small and far from the values of a typical acid mine drainage. In consequence, the static test may oversize the potential acidity of the tailing; whereas large columns may be saturated in water, displacing the oxygen and inhibiting the microbial activity necessary to catalyse mineral oxidation. (C) 2004 Elsevier B.V. All rights reserved.