A study was made, over a period of 29 weeks, of the interrelationship between calcium (Ca) and phosphorus (P) levels while monitoring those of magnesium (Mg) in ovine bone, soft tissues, faeces, urine and rumen fluid as well as blood plasma, where the P was determined as inorganic P (Pi) in order to determine whether the results reflected the dietary intake of P and Ca. Three groups of 6-month-old South African Mutton Merino x Merino wethers were fed 3 different rations containing, respectively, approximately the same amounts of crude protein (7, 28; 7, 67; 7, 53%) and Mg (0, 12; 0, 13; 0, 11 %), but increasing amounts of P (0, 05; 0, 27; 0, 63 %) and varying amounts of Ca (0, 22; 0,62; 0,20%). A comparison of the results for these Low, Medium and High phosphate (P) groups of animals showed that plasma Ca did not reflect dietary intake and, although the plasma Pi levels of the Low P group were low, those of the Medium and High groups rarely differed. Thus, in the assessment of the dietary intake of Ca and P by grazing sheep, plasma Ca and Pi levels must be viewed with extreme caution. On the other hand, the mean faecal Ca and P levels did reflect the dietary levels of these 2 elements. The Ca content of small faecal grab samples was closely correlated with the Ca content of 4-day pooled samples from the same animal (P< 0,01, r = 0,927). A similar finding was true for P (P<0,01, r = 0 ,963). Thus, taking rectal faeces from a number of sheep in a flock for individual or pooled analyses for Ca and P appears to be the best method for assessing the dietary intake of these elements. The Ca percentage in bone increased as dietary Ca increased but was influenced by the availability of P since the ratio of P to Ca in bone remained virtually constant. There was a significant correlation between the ash percentage and the midshaft width of the cortex of the femur (P< 0,01, r = 0,879). Significant negative correlations (P< 0,01) were also found between whole dry bone fat and ash percentage in cervical vertebrae (r = - 0,963), ribs (r= -0,903) and femurs (r= -0, 885). Thus ash analyses on whole dry bone will probably suffice to indicate the mineral status of the animals. Body composition estimated from tritiated water space was of little value as total body ash percentage was the reverse of the ash percentage actually found in the bones. Bone turnover as measured by urinary hydroxyproline excretion was significantly lower (P < 0,01) in the Low P group. The concentration of Ca and/or P in soft tissues did not reflect dietary intake. Similarly the Ca concentration in the case of urine and ruminal fluid did not reflect dietary intake whereas P concentrations did. Although the dietary intake of Ca and P was reflected in the mean daily urinary excretion, this measurement is unpractical in grazing animals. Although dietary Mg was almost equal in the 3 groups, faecal Mg was the lowest and plasma Mg the highest in the Low P group. Bone Mg was significantly lower (P< 0,01) in the Low P group than in the other 2 groups in respect of all 3 types of bone analysed.