The response of oilseed rape to applied nitrogen and phosphorus was investigated in 2 areas of the South East of South Australia. The nitrogen rates applied varied from 0 to 50 kg/ha, while phosphorus rates applied in the Mundulla area ranged from 0 to 20 kg/ha and 0 to 34 kg/ha in the Millicent area. At 9 sites in the Mundulla area, where soils were primarily sandy loam over clay, oilseed rape yields were increased significantly by applied nitrogen at 7 sites and by applied phosphorus at 2 sites. On heavy black clay and peat soils in the Millicent area, oilseed rape yields were increased significantly by applied nitrogen at 5 of the 12 sites and by applied phosphorus at 6 sites. Of the measured soil variables, anaerobic ammonium nitrogen measured in the top 10 cm best predicted responses in seed yield to applied nitrogen. The critical nutrient range was 45-65 mg/kg. The response of oilseed rape to applied nitrogen was highly correlated with past paddock use but varied between the 2 areas. From the prediction equations developed, it was concluded that, in the Mundulla area, a significant (P< 0.05) nitrogen response was likely if the oilseed rape followed 2 or more crops, the last a non-legume, but was unlikely if the oilseed rape followed 1 or more years of pasture. At Millicent, a nitrogen response was likely if the oilseed rape was grown as the sixth crop in a continuous cropping program, and was unlikely if it was the first or second crop in the rotation, provided the preceding crop was a non-legume. The response of oilseed rape to applied phosphorus was highly correlated to extractable soil phosphorus measured in the top 10 cm (Colwell). The critical nutrient range was 20-25 mg/kg for the sandy loam soils at Mundulla and 40-50 mg/kg for the black clay and peat soils at Millicent. Nitrogen applications significantly increased seed oil content at 6 sites, significantly decreased it at 2 sites and had no effect at 13 sites. Phosphorus significantly increased seed oil content at only 1 site, significantly decreased it at 3 sites, and had no effect at the other 17 sites.