The effect of extracellular Na+ removal and replacement with other cations on receptor-mediated arachidonate release in platelets was studied to investigate the role of Na+/H+ exchange in this process. Replacement with choline+, K+, N-methylglucamine+ (which abolished the thrombin-induced pHi rise) or Li+ (which allowed a normal thrombin-induced pHi rise) significantly decreased arachidonate release in response to all concentrations (threshold to supra-maximal) of thrombin and collagen. This inhibition was not reversed by NH4Cl (10 mM) addition, which raised the pHi in the absence of Na+, but, on the contrary, NH4Cl addition further decreased the extent of thrombin- and collagen-induced arachidonate release, as well as decreasing 'weak'-agonist (ADP, adrenaline)-induced release and granule secretion in platelet-rich plasma. No detectable pHi rises were seen with collagen (1-20 micrograms/ml) and ADP (10 microM) in bis-(carboxyethyl)carboxyfluorescein-loaded platelets. Inhibition of thrombin-induced pHi rises was seen with 0.5-5 microM-5-NN-ethylisopropylamiloride (EIPA), but at these concentrations EIPA had little effect on thrombin-induced arachidonate release. At higher concentrations such as those used in previous studies (20-50 microM), EIPA inhibited aggregation/release induced by collagen and ADP in Na+ buffer as well as in choline+ buffer (where there was no detectable exchanger activity), suggesting that these concentrations of EIPA exert 'non-specific' effects at the membrane level. The results suggest that (i) Na+/H+ exchange and pHi elevations are not only necessary, but are probably inhibitory, to receptor-mediated arachidonate release in platelets, (ii) inhibition of receptor-mediated release in the absence of Na+ is most likely due to the absent Na+ ion itself, and (iii) caution should be exercised in the use of compounds such as EIPA, which, apart from inhibiting the Na+/H+ exchanger, have other undesirable and misleading effects in platelets.