Measurements of net mineralization using a field incubation method were made over a full growing season (180 d). Soil cores, taken from cut swards which for many years had been previously grazed by cattle, were placed in jars in the field for successive incubation periods of 14 d. Acetylene was added to the incubation jars to inhibit nitrification in the soil cores and thereby prevent losses of N through denitrification. Net mineralization over 180d amounted to 415, 321 and 310kg N ha-: under grass/ clover, unfertilized grass and grass receiving 420 kg N ha-1 Y-1, respectively. At the start of the growing season, an index of potentially mineralizable N in the soil was estimated by a chemical extraction method, but this index was <50% of the estimates obtained by field incubation. The amount of N in herbage harvested regularly from the swards also under-estimated the supply of N from the soil, with apparent recoveries of 53, 82 and 74% and total yields of N of 240, 263 and 538 (kgN ha -1) from grass/clover, unfertilized grass and fertilized grass, respectively. Mineralization rates varied significantly with seasonal soil temperature fluctuations, but the incubation method was apparently less sensitive in relation to changes in soil water content. Rates of N-turnover (as % of total soil N) were highest under grass/clover (9%), but similar under fertilized and unfertilized grass swards (approximately 5%).