Available N was lowest (91'143 kg N/ha) following perennial grass'clover swards at the drier site where the annual legume content was lower, but perennial grass'clover swards produced larger soil N values (147'219 kg N/ha) at the higher rainfall site. Removal of the pasture in August'September compared with November in the year before cropping increased available N at the time of sowing by an average of 44% (51 kg N/ha) at the drier site and 43% (74 kg N/ha) at the wetter site. Incorporating perennial pasture species in swards was found to be advantageous in reducing nitrate leaching and preventing a decline in surface soil pH; however, available soil N to following crops could be lower if the annual legume content of perennial grass-based pastures declined due to competition from the perennial species.Soil mineral nitrogen (N) profiles during the growing season and changes in total soil N and available N after 3'4 years were examined under 9 different pasture swards containing annual legumes, lucerne (Medicago sativa L.), or one of 4 perennial grasses at 2 sites representative of the low and medium rainfall belt of south-eastern Australia. The effect of the presence of phalaris (Phalaris aquatica L.) or lucerne on the spatial variation in surface pH was also measured. The 9 pastures were subterranean clover (Trifolium subterraneum L.), subterranean clover with annual weeds, yellow serradella (Ornithopus compressus L.), lucerne, phalaris, cocksfoot (Dactylis glomerata L.), lovegrass (Eragrostis curvula (Schrader) Nees), wallaby grass (Austrodanthonia richardsonii (Cashm.) H.P. Linder), and a mixture of lucerne, phalaris, and cocksfoot. All the perennial treatments were sown with subterranean clover. Available mineral N values in the surface 0.10 m of soil following summer rainfall were substantially higher in pure subterranean clover or serradella (Ornithopus compressus L.) swards (24'50 'g N/g) than those containing a mixture of subterranean clover and perennials (9'20 'g N/g). Apparent leaching of soil nitrate down the profile during winter was greatest in annual pasture treatments and least in swards containing perennials. Soil pH(CaCl2) at the 0'0.10 m depth varied with proximity to perennial plants and was significantly higher (+0.2'1.1 pH units) near the base of perennial plants than in gaps between the perennials or in annual-only swards. Available mineral N to 1.0 m before cropping at the end of the pasture phase was highest following subterranean clover (175'344 kg N/ha) and serradella (202'316 kg N/ha) at both sites.