1. The binding sites labelled by [35S]-adenosine 5'-O-[3-thiotriphosphate]([35S]-ATP gamma S) at 4 degrees C in rat vas deferens membranes were studied and compared to the sites labelled by [3H]-alpha,beta-methylene ATP ([3H]-alpha beta meATP) to ascertain whether [35S]-ATP gamma S can be used to label the P2x purinoceptor. 2. In the presence of 4 mM CaCl2, the binding of 0.2 nM [35S]-ATP gamma S to vas deferens membranes was increased 3.4 fold, when compared to studies performed in the absence of calcium. However, binding did not appear to be solely to P2x purinoceptors since [35S]-ATP gamma S labelled a heterogeneous population of sites and about 72% of the sites possessed high affinity (pIC50 = 7.5) for guanosine 5'-O-[3-thiotriphosphate] (GTP gamma S). Even in the presence of 1 microM GTP gamma S, to occlude the sites with high affinity for GTP gamma S, the binding of [35S]-ATP gamma S was heterogeneous and since there was also evidence of extensive metabolism of ATP in the presence of calcium, the binding of [35S]-ATP gamma S under these conditions was not studied further. 3. In the absence of calcium ions, [35S]-ATP gamma S bound to a single population of sites (pKD = 9.23; Bmax = 4270 fmol mg-1 protein). Binding reached steady state within 3 h (t1/2 = 38 min), was stable for a further 4 h and was readily reversible upon addition of 10 microM unlabelled ATP gamma S (t1/2 = 45 min). In competition studies the binding of 0.2 nM [35S]-ATP gamma S was inhibited by a number of P2x purinoceptor agonists and antagonists, but not by adenosine receptor agonists, staurosporine (1 microM) or several ATPase inhibitors. The rank order of agonist affinity estimates (pIC50 values) in competing for the [35S]-ATP gamma S binding sites was: ATP (9.01), 2-methylthio- ATP (8.79), ATP gamma S (8.73), alpha beta meATP (7.57), ADP (7.24), beta, gamma-methylene ATP (7.18), L-beta, gamma-methylene ATP (5.83), alpha, beta-methylene ADP (4.36). 4. Affinity estimates (pIC50 values) for the P2x purinoceptor antagonists, suramin (5.20), pyridoxalphosphate-6-azophenyl-2',4'-disulphonic acid (4.23), pyridoxal 5-phosphate (3.42), cibacron blue (5.70) and Evan's blue (5.79) were broadly similar to those obtained at the [3H]-alpha beta meATP binding sites in vas deferens. However, ATP, 2-methylthio-ATP, ATP gamma S and ADP displayed 17-512 fold higher affinity for the [35S]-ATP gamma S, than for the [3H]-alpha beta meATP binding sites, whereas alpha beta meATP and L-beta, gamma-methylene ATP displayed 5 and 28 fold, respectively, higher affinity for the [3H]-alpha beta meATP than for the [35S]-ATP gamma S binding sites. 5. The differences in agonist affinity for the [35S]-ATP gamma S and [3H]-alpha beta meATP binding sites probably reflect the fact that the former sites were labelled in the absence of calcium, while the latter sites were labelled in its presence. This could differentially affect ionisation state and/or metabolism of the nucleotides when using the two radioligands. Since affinity estimates for ATP, 2-methylthio-ATP, ATP gamma S, alpha beta meATP and L-beta, gamma-methylene ATP were different when calcium ions were omitted in studies using [3H]-alpha beta meATP but similar to the affinity estimates obtained at the [35S]-ATP gamma S binding sites labelled in the absence of calcium, it is likely that [35S]-ATP gamma S and [3H]-alpha beta meATP label the same sites in rat vas deferens. 6. We conclude that, in the absence of divalent cations, [35S]-ATP gamma S labels P2x purinoceptors in rat vas deferens and as such may represent a new, high specific activity, radioligand for the study of such receptors.