The purpose of the current work was to study phosphorus (P) metabolism in growing sheep supplemented with different levels of dicalcium phosphate using an extant mathematical model. Twelve male non-castrated Santa Inês sheep, weighing 23 (±2·2) kg, received a basal diet unsupplemented or supplemented with dicalcium phosphate to provide 1·5, 3·0, 4·5 g of P/animal per day (treatments T1 to T4, respectively). After 3 weeks adaptation, 7·4 MBq of 32P was injected into the jugular vein of each animal. Samples of blood, faeces and urine were collected every day during a 7-day period and thereafter the animals were sacrificed and samples from liver, kidney, heart, muscle and bone were collected for specific activity and inorganic P determinations. The flows between gut and plasma were similar for each treatment except for T1, which showed the lowest values for both flows (P<0·05). The amount of P accreted to soft tissue (F42) was different among treatments, however net tissue retention was similar for all treatments. Total P retained was highest for T4 and lowest as well as negative for T1 and T2. Phosphorus accreted to bone (F32) was different among treatments and contributed to the different net bone retentions. The highest value of F32 was reached by animals on T4, whilst the lowest values were found for animals on T1. Despite having the highest value of F32, it should be noted that animals on T4 excreted the most P in faeces. Considering concerns about environmental P pollution, it is important to be aware that the treatment which provided the highest value for net bone P retention and for F42 also led to the highest value of P excreted in faeces. Therefore, the current study suggests that T3 provided the best P level for this category of animal since P accreted to bone and tissue indicated that P absorption was adequate to attend to P requirements.