Abstract.The decline in topsoil (0–0.1 m) pH (CaCl2) over 11 years (1979–90) was measured in a rotation, tillage, stubble and nitrogen fertilizer experiment on a Chromic Luvisol at Wagga Wagga in S.E. Australia. The rotations consisted of annual wheat cropping (WW) with and without nitrogen fertilizer (100 kg N/ha/year), alternating lupin‐wheat (LW) and subterranean clover‐wheat (CW). The initial mean pH at the site was 4.9 and the experiment was preceded by subterranean clover‐based pasture for most of the previous 19 years. An initial rapid decline in soil pH under all treatments over the first 8–9 years was followed by a 2–3 year period when no further decline was detected. The annual rate of pH decline over the first 8–9 years varied from 0.06 for WW to 0.09 units for WW with added N fertilizer. Apparent steady‐state for WW after 11 years was approximately 0.5 pH units higher than for WW with added N fertilizer. There was no difference between CW and LW in the rate of decline or in the apparent steady‐state reached. Six years’ stubble burning in a LW rotation promoted a slightly higher pH than where stubble was retained. However, there was no significant effect of tillage in either LW or CW rotations. By 1990 the addition of N fertilizer to WW had increased the concentration of exchangeable aluminium by 100% and of manganese by 24%. The inclusion of a legume in the rotation increased the concentration of aluminium but did not affect manganese. However, burning stubble in the LW rotation slightly decreased man