The Na/K pump in human red blood cells that normally exchanges 3 Nai for 2 Ko is known to continue to transport Na in a ouabain-sensitive and ATP-dependent manner when the medium is made free of both Nao and Ko. Although this Na efflux is called "uncoupled" because of removal of ions to exchange with, the efflux has been shown to be comprised of a coefflux with cellular anions. The work described in this paper presents a new mode of operation of uncoupled Na efflux. This new mode not only depends upon the combined presence of ADP and intracellular orthophosphate (P(i))i but the Na efflux that is stimulated to occur is coeffluxed with (P(i))i. These studies were carried out with DIDS-treated resealed red cell ghosts, suspended in buffered (NMG)2SO4, that were made to contain, in addition to other constituents, varying concentrations of ADP and P(i) together with Na2 SO4, MgSO4 and hexokinase. While neither ADP nor P(i) was effective alone, ouabain-sensitive uncoupled Na efflux, (measured with 22Na) could be activated by [ADP+P(i)] where the K0.5 for ADP in the presence of 10 mmol (P(i))i/liter ghosts was 100-200 mumol/liter ghosts and the K0.5 for (P(i))i, in the presence of 500 mumol ADP/liter ghosts was 3-4 mmol/liter ghosts. [ADP+P(i)] activation of this Na efflux could be inhibited by as little as 2 mumol ATP/liter ghosts but the inhibition could be relieved by the addition of 50 mM glucose, given entrapped hexokinase. While ouabain-sensitive Na efflux was found to be coeffluxed with P(i) (measured with entrapped [32P]H3PO4), this was not so for SO4 (measured with 35SO4). The stoichiometry of Na to P(i) efflux was found to be approximately 2 to 1. Na efflux as well as (P(i))i efflux were both inhibited by 10 mM Nao (K0.5 approximately equal to 4 mM). But, whereas 20 mM Ko (K0.5 approximately equal to 6 mM) inhibited the efflux of (P(i))i, as would be expected from previous work, Na efflux was actually increased. When Ko influx was measured in this situation there was a 1 for 1 exchange of Nai for Ko, that is, of course, downhill with respect to the gradient of each ion. Surprisingly AsO4 was unable to replace P(i) for activation of Na efflux but Na efflux could be inhibited by vanadate and oligomycin. In terms of mechanism, it is likely that ADP acts to promote the formation of the phosphoenzyme (EP) by (P(i))i that would otherwise be inhibited by Nai.(ABSTRACT TRUNCATED AT 400 WORDS)