The specific activity of the gamma-32P position of ATP was measured in various tissue preparations by two methods. One employed HPLC and the enzymatic conversion of ATP to glucose 6-phosphate and ADP. The other was based on the phosphorylation of histone by catalytic subunit of cAMP-dependent protein kinase (Hawkins, P.T., Michell, R.H. and Kirk, C.J. (1983) Biochem. J. 210, 717-720). The HPLC method also allowed the incorporation of 32P into the (alpha + beta)-positions of ATP to be determined. In rat epididymal fat-pad pieces and fat-cell preparations the specific activity of [gamma-32P]ATP attained a steady-state value after 1-2 h incubation in medium containing 0.2 mM [32P]phosphate. Addition of insulin or the beta-agonist isoprenaline increased this value by 5-10% within 15 min. Under these conditions the steady-state specific activity of [gamma-32P]ATP was 30-40% of the initial specific activity of the medium [32P]phosphate. However, if allowance was made for the change in medium phosphate specific activity during incubations the equilibration of the gamma-phosphate position of ATP with medium phosphate was greater than 80% in both preparations. The change in medium phosphate specific activity was a combination of the expected equilibration of [32P]phosphate with exchangeable intracellular phosphate pools plus the net release of substantial amounts of tissue phosphate. At external phosphate concentrations of less than 0.6 mM the loss of tissue phosphate to the medium was the major factor in the change in medium phosphate specific activity. It is concluded that little advantage is gained in employing external phosphate concentrations of less than 0.6 mM in experiments concerned with the incorporation of phosphate into proteins and other intracellular constituents. Indeed, a low external phosphate concentration may cause depletion of important intracellular phosphorus-containing components.