Simmentaler cows grazing Cymbopogon-Themeda veld in the western Highveld region of South Africa (1985/861989/90) received three levels of phosphorus (P) supplementation. Dicalcium phosphate content of the supplement was manipulated to give intakes of 8 (CS), 4 (HS) or 0 (ZS) g P/cow/day during summer. All cows received supplemental P (10 g/cow/day) during winter (1986-1989). In 1990, supplemental P (10 (CW), 5 (HW) or 0 (ZW) g P/cow/day) was given in a winter maintenance supplement (protein, energy and minerals). Rainfall was above average during the trial period. Reproductive performance was not influenced by P supplementation. Mean livemass of the CS group was greatest (p < 0.01), and that of HS was greater (p < 0.01) than ZS at the end of summer. Both CS and HS had better (p < 0.01) condition scores than ZS. Bone P content was lowest and indicative of deficiency (p < 0.01) in ZS cows (112.6 vs 141.9 and 130.8 mg P/cm3 bone for CS and HS respectively). Calf performance was not affected (p > 0.05) by P supplementation. Cow mass was affected by winter P supplementation (p < 0.05). Both CW and HW displayed improved (p < 0.01) condition scores and higher (p < 0.01) bone P content than ZW. Fatalities (4) occurred in ZW due to P deficiency (74.5 mg P/cm3 bone), and deficiency symptoms were manifested in the entire group. Both summer and winter veld is deficient in P, which makes continuous supplementation a recommended practice in this area.