1. Ten castrated young adult sheep were brought indoors and given lucerne chaff (four animals with a phosphorus intake of 2 . 40 g/d) and fresh Ruanui ryegrass (six animals with a P intake of 3 . 72 g/d) to determine various kinetic indices of their P metabolism. 2. The results obtained from combined nutritional balance and 32P radioactivity measurements were analysed using the simulation, analysis and modelling computer (SAAM) program. Since the plasma P specific activity curve was described by at least four exponential terms, a four compartmental model (M1, M2, M3 and M4) was used to describe the experimental values. 3. The mean mass of the total exchangeable P pool (MT) for the lucerne chaff-fed and Ruanui ryegrass-fed sheep was 25 . 5 and 28 . 8 g respectively, while the total P transport (VT) to the exchangeable pool (MT) was markedly lower for sheep given lucerne chaff compared to those given Ruanui ryegrass (54 . 1 v. 111 . 3 mg/kg body-weight per d). Likewise the rate of P absorption (Va), faecal endogenous loss of P (Vf), urinary P excretion (Vu) as well as the movement of P to (Vo+) and from (Vo-) the skeleton and soft tissue was also lower in the lucerne chaff-fed sheep. 4. The availability of P, that is the proportion of ingested P which is absorbed, was 0 . 56 and 0 . 62 for the sheep given chaff and fresh herbage respectively. 5. The P metabolism of one sheep (sheep no. 30) differed from the other sheep in that its plasma inorganic P concentration was higher (95 v. 54 mg/l). This was associated with a larger first compartment (M1), a more rapid P transport between M1, and M2, an increase in (Vo) and a marked increase in (Vu).