A study has been made of the mechanism by which the injection of ATPNa2 stimulates the ouabain-insensitive Na efflux in fibers from the barnacle, Balanus nubilus. The results of this study are as follows: ATPNa2 is found to be a more potent effector of the Na efflux in unpoisoned fibers than ATPMg on an equimolar basis, but not more potent than ADPNa2. In ouabain-poisoned fibers ATPNa2 and ATPMg are equipotent but the former is more potent than ADPNa2. The magnitude of the response to ATPNa2 injection into ouabain-poisoned fibers depends on: (i) the ouabain concentration used; (ii) the concentration of ATPNa2 injected, and (iii) the external Ca2+ concentration. Ouabain is without effect when it is applied at the time of ATPNa2 injection. Responsiveness to ouabain, however, is found to return if the glycoside is applied after complete decay of the response to ATP. Under these conditions, the effect of ouabain in fibers injected with ATPNa2 is significantly less than in fibers injected with ATPMg. Preinjection of EGTA in high concentrations fails to reduce the size of the response to ATPNa2 injection. Injection of Mg2+ following peak stimulation by ATP almost completely reverses the response. The response to Mg2+ is concentration-dependent. Ryanodine but not neomycin reduces the response to ATP. ATP gamma S is not as effective as ATPNa2. Nor is AMP-PNP consistently as effective as ATPNa2. Collectively, these results support the hypothesis that the response of the Na efflux to ATPNa2 injection involves the operation of the putative Na(+)-Ca2+ exchanger in the reverse mode and that a raised Cai2+ is not an absolute requirement. They also strongly suggest that two other governing factors are the Na+ gradient across the sarcolemma and the myoplasmic pMg. Mg2+ seems to act as an inhibitor.