To estimate the behaviour of metals in soils during pH change, polluted soil from an industrial area was subjected to sulphuric acid and calcium hydroxide additions. The behaviour of metals was studied by speciation techniques with separation into five fractions, namely: exchangeable, linked to carbonates, organic matter + sulphides or Fe‐Mn oxides and rest (associated to primary materials). As extractants, ammonium acetate, acetic acid, hydrogen peroxide and hydoxylamine were successively used. Determined metals include macro‐ (Na, K, Ca, Mg, Fe) and microconstituents (Cu, Pb, Mn, Zn) measured by AAS directly from extractant solutions and after HNO3—FH—HCIO4 digestion for total content. Analytical data are discussed as percentages of each metal in the different fractions at different soil pH values. The following results were obtained: (i) some macroconstituents (Mg, K, Na) are not influenced by acid‐base treatments of soil, (ii) calcium is probably retained as gypsum by soil treated with sulphuric acid, but an increase of its exchangeable amount was observed, (iii) most of Fe remains in the residual fraction, but the amount available for plants increases at low soil pH, and, (iv) the studied microconstituents become more soluble and presumably available for plants at lower soil pH values. As a whole, no differences were seem for fractions associated to Fe—Mn oxides or organic matter + sulphides, a decrease in the total amount of metals as carbonates and minima at intermediate pH values for exchangeable and metals linked to residual fraction were observed. This behaviour is further discussed in the sense that soil is a natural buffer, whose pH changes may be explained by conventional acid‐base equilibria. From the experimental data it was possible to estimate the acidity constant, which is probably related to the presence of clays in the soil studied.